article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
this work was done together with georges ripka and wojciech broniowski .
solitons corresponding to baryons have been found in several chiral quark models .
many of these solutions turn out to be unstable against collapse unless additional constraints are introduced in the model .
the well known examples are the linear njl model with proper time regularization @xcite and the linear @xmath0-model with sea quarks @xcite . even in the linear @xmath0-model with only valence
quarks the energy of the soliton becomes too low for any choice of model parameters if one goes beyond the mean field approximation .
in all these models the instability occurs because it is energetically favorable for the chiral field to acquire arbitrary ( or very ) high gradients .
this suggests that cutting off high momenta in the interaction may prevent the collapse and stabilize the soliton .
a simple sharp cut - off does not yield a stable solution while a smooth behavior of the regulator ( usually interpreted as a @xmath1-dependent quark mass ) can indeed lead to solitons which are stable against the decay into free quarks as well as against collapse .
such a regularization has a physical justification in qcd calculations of the quark propagation in an instanton liquid which predict a non - local effective interaction between quarks with a 4-momentum cut - off @xmath2 mev @xcite .
further physical implications of the non - local regularization are discussed in the contributions to this workshop by george ripka and wojciech broniowski @xcite .
the non - local regularization of the quark - quark interaction can be implemented in the njl type models by replacing the contact term @xmath3 , @xmath4 by a non - local form .
usually one introduces a regulator @xmath5 diagonal in 4-momentum space such that @xmath6 .
the qcd derivation of the quark propagation in a dilute instanton gas predicts the following functional dependence for @xmath7 @xcite : @xmath8 where @xmath9 is the instanton size of the order @xmath10 . as we shall see in the following it
is necessary to analytically continue the regulator to negative @xmath11 in order to be able to treat the valence orbit .
this is not possible with the form ( [ instanton ] ) since it has a cut along the negative real axis starting at @xmath12 .
we use instead a gaussian shape of the regulator : @xmath13 or a `` monopole '' shape : @xmath14 which has the proper behavior for large @xmath11 where one gluon exchange dominates .
the expression for the energy of the soliton and the self - consistency equations can be derived from the bosonized euclidean action @xmath15 where @xmath16 and @xmath17 are the chiral fields and are the dynamical variables of the system .
the main difficulty is the presence of time in the regulator . in order to evaluate the trace in ( [ action ] ) it is convenient to introduce energy dependent basis states , which are solutions of the dirac equation : @xmath18 with @xmath19 from ( [ action ] ) the following expression for a stationary configuration can be derived @xcite : @xmath20 note that when no regulator ( i.e. @xmath21 ) or a time - independent regulator is used , the energies @xmath22 are independent of @xmath23 and the integration can be carried out using the cauchy theorem . closing the contour from below yields the well known expression for the energy of the dirac sea : @xmath24 .
( note that the energies of occupied orbits lie on the negative imaginary @xmath23-axis . )
when the soliton describes a baryon , the energy of three valence quarks is added to the energy of the dirac sea .
the same result can be formally obtained by deforming the contour in ( [ energy ] ) in such a way as to encircle the valence orbit ( for detailed discussion on this point see wojciech broniowski contribution to this workshop ) .
such a prescription gives the expected result provided the orbits do not depend on @xmath23 .
however , when the regulator depends on time ( or @xmath23 ) , this may not lead to the correct result since the regulator generates additional poles scattered in the whole complex @xmath23-plane .
it may still work well for an isolated pole on the positive imaginary axis close to 0 as is the case of the @xmath25 orbit in the soliton with the hedgehog form of the background chiral field @xcite .
this pole can then be treated separately , yielding the valence contribution to the soliton energy @xmath26 , where the energy of the valence orbit is determined from @xmath27 the soliton energy can now be written as : @xmath28 the sea contribution is @xmath29\ ] ] with @xmath30 and is evaluated by direct numerical integration along the real @xmath23-axis . the term @xmath31 is given by the last integral in ( [ energy ] ) ( with the integrand @xmath32 ) .
the above prescription is further supported by the fact that it gives an exact result for the baryon number , which can be expressed as @xcite : @xmath33 the self - consistent equations derived from ( [ action ] ) take the form ( the hedgehog ansatz , @xmath34 , for the pion field is assumed ) : @xmath35\ ; , \label{sceq}\end{aligned}\ ] ] where @xmath36 and @xmath37 is the residue of the valence pole .
a necessary condition for a stable soliton configuration is that the energy ( [ energy ] ) is lower than the energy of three free quarks .
when the regulator depends on time , the free quark mass , @xmath38 , is not simply the vacuum value of the chiral field , @xmath39 , but is determined by the position of the pole of the quark propagator in the vacuum @xcite , _
i.e. _ it corresponds to the solution of @xmath40 .
the solution for real @xmath11 exists only below a critical value of @xmath39 ( see figure [ figep ] ) ; above this point no stable free quarks exist .
however , a stable solution can always be found beyond this point provided the quarks dress in a spatially non - uniform background chiral field . ' '' '' height 0pt depth 0pt
the model ( [ action ] ) possesses 4 parameters : the vacuum value of the chiral field @xmath39 , the current quark mass @xmath41 , the coupling constant @xmath42 and the cut - off parameter @xmath43 .
the coupling constant @xmath42 is fixed from the stationarity condition in the vacuum , @xmath43 is fitted to the pion decay constant @xmath44 , and @xmath41 to @xmath45 .
we are left with one free parameter @xmath39 . in this section
we analyze how different shapes of the regulator affect the result .
we compare the gaussian ( [ gaussian ] ) , the monopole ( [ pole ] ) and a modified version of the instanton ( [ instanton ] ) regulator .
as we have mentioned it is not possible to continue ( [ instanton ] ) to negative @xmath11 , which is needed in order to obtain the valence contribution .
we have instead introduced a form which is identical to the instanton regulator for @xmath46 while its behavior for @xmath47 is replaced by a real function which approximately follows the real part of ( [ instanton ] ) for small negative @xmath11 ( `` extended instanton '' ) .
the gaussian and the monopole shapes lead to practically identical results .
we have also tried other shapes and found very similar results .
the reason is that several properties ( including the integral that determines @xmath44 ) depend mostly on the behavior of the regulator for small value of @xmath11 .
if this behavior is @xmath48 then @xmath49 is almost uniquely determined by the value of @xmath44 .
hence , all shapes with this type of behavior lead to very similar results .
the situation is quite different if @xmath50 , _ e.g. _ for regulators that depend only on @xmath51 .
we do not find stable solutions for such regulators ; the energy of the dirac sea is always higher than the gain due to the lowering of the valence energy .
the form ( [ instanton ] ) has neither of the above behaviors for @xmath52 but behaves as @xmath53 .
it is therefore interesting to check whether stable solitons can also be obtained for such a particular form .
we indeed find solitons with very similar properties to those obtained using ( [ gaussian ] ) or ( [ pole ] ) . since @xmath44 sets the scale , it is also interesting to study how a higher value for @xmath44 would affect the results .
.properties of the self - consistent soliton solutions for different shapes of the regulator , `` extanton '' stands for the extended instanton regulator with @xmath54 , gaussian@xmath55 means a gaussian regulator with @xmath43 fitted to @xmath56 mev , @xmath57 is the one - flavor quark condensate . [
cols="<,^,^,^,^,^,^,^,^,^ " , ] ' '' '' height 0pt depth 0pt ' '' '' height 0pt depth 0pt a ) ' '' '' height 0pt depth 0pt ' '' '' height 0pt depth 0pt b ) [ com ]
the calculation can be made much simpler and the problem of analytic continuation to negative @xmath11 avoided if we take a regulator that does not depend on time .
then the states do not depend on @xmath23 and the integration over @xmath23 in the above expressions can be carried out analytically .
all quantities still remain finite and the soliton remains stable against collapse .
the parameters used are the same as for the 4-momentum regulator .
however , the explicit breaking of lorentz invariance leads to serious problems which we discuss in this section .
the energy of the soliton reduces to @xmath58 and the self - consistency equation to @xmath59\;. \nonumber\end{aligned}\ ] ] the simplest form of the regulator is a sharp cut - off limiting the 3-momenta to @xmath60 .
since one usually uses basis states with good @xmath61 , this is equivalent to restricting the hilbert space .
such a form is very easy to implement since the regulator does not appear explicitly in the calculation .
unfortunately , no stable soliton solution is found in this case ; the contribution of the sea quarks is always larger than the gain due to the lowering of the valence energy .
stable solutions can be obtained if one takes a smooth form of the regulator , e.g. a gaussian form @xmath62 .
the results are displayed in figure [ fige3d ] .
the threshold value of @xmath39 below which the soliton does not exist is somewhat larger than for the 4-momentum gaussian regulator .
one notices a rather striking feature that the soliton does not exist _ beyond _ a certain value of @xmath39 .
the reason for such behavior is the following : the energy of a free quark can be written as @xmath63 . for sufficiently small @xmath64
the minimum of @xmath65 is not at @xmath66 but rather at some @xmath67 , and the energy of a free quark becomes smaller than @xmath39 . increasing @xmath39 the cut - off @xmath43 decreases and at a certain value
it becomes energetically favorable for the quarks in the soliton to acquire sufficiently high momenta and leave the soliton .
this happens when @xmath68 .
such an unphysical behavior is a clear consequence of breaking the lorentz invariance in the interaction .
the value of @xmath69 can be easily determined if we choose a gaussian regulator and @xmath70 . for the values of @xmath43 below @xmath71
the minimum of @xmath72 occurs for @xmath73 and the free quark energy at the minimum is : @xmath74 . for @xmath75 this value
is slightly modified : @xmath76 ^ 2\;.\ ] ] figure [ fige3d ] shows that the soliton exists for @xmath77 mev . at @xmath78
mev the cut - off @xmath43 reaches the critical value and the lowest free quark state has @xmath79 . nonetheless , the soliton remains bound since @xmath80 . at @xmath81
mev the condition is no longer fulfilled and no bound solution exists beyond this point . between these two values
the energy of the soliton differs only little from the energy of three free quarks .
the soliton is weakly bound ; the chiral field stays close to its vacuum value while the soliton radius is large . ' '' '' height 0pt depth 0pt
it is well known that the njl model can be transformed into the form of an equivalent linear @xmath0-model .
this transformation is explained in @xcite , chapter 5 .
if one assumes a sharp cut - off and sufficiently large @xmath43 ( compared to @xmath39 ) the lagrangian density of the equivalent @xmath0-model takes the familiar form @xmath82 the fields @xmath0 and @xmath83 are related to the two components of the chiral field of the bosonized njl model as @xmath84 , @xmath85 , where @xmath86 .
the parameter @xmath87 and the mass of the @xmath0 meson are related to the parameters of the njl model by here is a factor 2 smaller than the one in @xcite . ]
@xmath88 , @xmath89 and @xmath90 the question remains whether the above assumptions are met in the model described in section 2 .
is close to 1 while the smooth form of the regulator may and does generate additional terms in the lagrangian ( [ elsm - fam ] ) and modifies the values of the parameters .
a work that will take into account the additional terms is in progress .
the conclusions in this section remain valid but the results should be considered only as qualitative . ]
we assume that the lagrangian ( [ elsm - fam ] ) describes the sea quarks while the valence quarks are treated separately .
this is in the spirit of the approaches used in @xcite .
however , in all these approaches the valence orbit is not regularized and as we shall see in this section this brings a qualitative difference with respect to the situation where it is regularized . if we agree that a regulator in the quark - quark or , equivalently , quark - chiral field interaction has a well grounded physical origin , the valence orbit should be regularized in the same way as the orbits in the dirac sea .
the valence orbit is determined as in ( [ eval ] ) .
the soliton energy then becomes ( @xmath91 ) : @xmath92 + g \left [ \sigma j^\sigma + \pi_a j_a^\pi \right ] \right ) + e_{s.i.}\;,\ ] ] where the meson self - interaction term ( the `` mexican hat '' ) is given by @xmath93 here we have introduced the source terms which explicitly contain the regulator ( see ( [ sceq ] ) ) : @xmath94 the mean - field solution is obtained by solving the self - consistency equations ( the hedgehog ansatz @xmath95 and @xmath96 is assumed ) : @xmath97\ ; , \nonumber\\ & & \left({\d^2\over\d r^2 } + { 2\over r}\,{\d\over\d r } - { 2\over r^2 } - m_\pi^2\right)\pi(r ) = j^\pi_a(\vec{r})\hat{r}_a + \lambda^2\left [ { \sigma'(r)}^2 + 2f_\pi{\sigma'(r ) } + \pi(r)^2\right]\pi(r)\;. \label{sigmaeq}\end{aligned}\ ] ] in figures [ figelsm ] and [ mflsm ] we compare the properties of the soliton when the valence orbit is regularized and when it is not .
close to the threshold the two solutions do not differ much as it should be since here the cut - off @xmath43 is relatively large . for higher values of @xmath39 ( or @xmath98 )
the behavior is qualitatively quite different ; in particular , the energy of the unregularized valence orbit soon becomes negative which never happens for the regularized one .
let us finally mention why the regularization is important also in the linear @xmath0-model with only valence quarks . even if there is no instability as in the regularized linear @xmath0-model - propagator @xcite .
] , one encounters serious difficulties when going beyond the mean field approximation . in the @xmath0-model it is possible to use the peierls - yoccoz projection of good linear momentum , spin and isospin in order to obtain physical nucleon states .
it turns out that the energy of the soliton after projection is strongly reduced already when the mean - field solution is used for the chiral fields @xcite .
if in addition one allows a variation of the chiral field profiles , one obtains a solution with an energy considerably lower than the nucleon mass .
the energy gain is mostly due to a strongly localized chiral field which lowers the valence energy .
such a strong localization is not allowed when the regulator is used since it puts a physical limit on the gradients of the field ( through the source term ( [ source ] ) ) . | we describe how the non - local regularization can be implemented in the calculation of solitons in the nambu jona - lasinio model as well as in the equivalent linear @xmath0-model .
we investigate different forms of regulators and show that the 3-momentum cut - off leads to serious conceptual difficulties . |
i would first like to thank my supervisors hani mehrpouyan and alexandre graell i amat for their support and guidance during the whole thesis process .
i highly appreciate their attitude towards me .
they have taught me how an academic should approach the problems in the field of research .
they have also trusted me and given me a lot of freedom .
they respect my independent and somehow arrogant way of performing research .
i would also like to thank my friends in communication engineering masters programme for the discussions and shared ideas .
i want to also thank to the employees at ericsson for showing me how things are done in industry in a short time .
thanks also to my family and friends in turkey for supporting me through all these years of studying .
i have to thank my swedish family , jan and ann - charlotte fonselius , for creating such a nice environment for me at the house which i share with my beloved friend kiryl kustanovich .
last but not least , i want to thank olric . i would not be able to finish this thesis without him .
arif onder isikman , gteborg
in recent years the demand for high bandwidth data services has increased with the evolution of the third generation ( 3 g ) and fourth generation ( 4 g ) cellular networks @xcite .
rapid escalation in the use of bandwidth hungry devices also increases the throughput requirements of the base station ( bs ) , base station controller ( bsc ) and master switching center ( msc ) , which are the fundamental components of a cellular network .
the user connects to the network through the bs .
each bs is connected to a bsc via a wired or a wireless link .
the bsc routes the data from the bs to the msc and controls the functionality of the bs .
the msc holds all the network information and controls all calls and data management functionalities .
in other words , the msc is the brain of any cellular network .
the portion of a wireless mobile network from the bs to the msc is called as _
backhaul network_. the backhaul links serves the medium to carry traffic from the bs to the msc via the bsc .
the point - to - point microwave radio links are commonly used in backhaul networks .
they are cost efficient and can be deployed rapidly .
microwave radio transmission is operated at certain frequency bands .
lower bands such as 7 , 18 , 23 and 35ghz have better radio propagation characteristics . on the other hand ,
these frequency bands fail to provide sufficient bandwidth since the spectrum is mostly allocated . with the release of the e - band , 10ghz of bandwith in the spectrum at 70ghz ( 71 - 76ghz ) and 80ghz ( 81 - 86ghz )
have been made available for point - to - point microwave links . to meet high data rate requirements point - to - point microwave systems
are equipped with multiple transmit and multiple receive antennas . _ line - of - sight _
( los ) _ multi - input multi - output _ ( mimo )
systems are effectively used for backhaul networking @xcite .
local oscillators are utilized to carry the baseband signal to the operating band . due to the hardware limitations ,
every oscillator suffers from an instability of its phase , resulting in phase noise @xcite .
phase noise can dramatically limit the performance of a wireless communication system if left unaddressed @xcite .
phase noise interacts with the transmitted symbols both at the transmitter and the receiver side in a non - linear manner and significantly distorts the received signal .
digital signal processing algorithms need to be employed to achieve synchronous transmission in the presence of phase noise .
several algorithms are proposed for _ single - input single - output _ ( siso ) systems to mitigate the effect of time varying phase noise @xcite . in the case of los - mimo systems ,
each transmit and receive antenna is equipped with a different oscillator since the antennas are placed far apart .
similarly , in the case of multi - user mimo systems or _
space division multiple access _ ( sdma ) systems independent oscillators are used by different users to transmit their data to common receiver @xcite . as a result ,
a single oscillator can not be employed and phase noise compensation algorithms proposed for siso systems are not directly applicable to mimo systems . achieving channel capacity
was seen far from reality until two decades ago . the introduction of turbo codes @xcite and the rediscovery of _ low - density - parity - check ( ldpc ) _
codes @xcite has demonstrated the power of the iterative processing paradigm in improving the performance of communication systems and in operating close to the theoretical limits .
subsequently , the iterative coding structure has been applied to facilitate and improve many functions including synchronization .
parameter estimation can be performed jointly with data detection in an iterative fashion .
it is well - known that the application of turbo codes and lpdc codes improves the data detection process at the receiver , which in turn can be applied to improve the performance of decision - directed estimators .
the improved estimation and tracking accuracy allows for more accurate compensation of impairments such as time varying phase noise at the receiver which can also improve data detection .
thus , by jointly performing data detection and estimation , the performance of wireless communication systems can be significantly improved .
this approach , known as turbo synchronization " , was initially proposed in @xcite and has since been formalized in @xcite with the use of the _ expectation - maximization ( em ) _
framework @xcite . in @xcite ,
different frameworks for turbo synchronization based on the gradient method and the sum - product algorithms are studied .
this work is extended to the problem of estimation of time varying phase noise for siso systems in @xcite . in @xcite ,
based on the assumption of small phase noise values within each block and removing the data dependency from the observed signal , the tracking is carried out via a modified em - based algorithm that applies a soft decision - directed linearized kalman smoother . in addition , to enhance phase noise tracking performance for very high phase noise variances , @xcite proposes to employ a maximum - likelihood ( ml ) estimator in conjunction with a kalman smoother , labeled as ( ks - mla ) .
a soft decision - directed extended kalman filter - smoother ( eks ) is also suggested to provide phase noise estimation .
however , the performance of the ks - mla degrades with increasing block length .
more importantly , the linearization applied in @xcite is not applicable to mimo systems and the estimation performance of the proposed tracking algorithm is not investigated .
mimo technology allows communication systems to more efficiently use the available spectrum @xcite,@xcite . _
bit - interleaved - coded - modulation ( bicm ) _ is one of the popular schemes that enables communication systems to fully exploit the spectrum efficiency promised by mimo technology @xcite .
however , the performance of mimo systems degrades dramatically in the presence of synchronization errors .
code - aided synchronization based on the em framework for joint channel estimation , frequency and time synchronization for a bicm - mimo system is proposed in @xcite .
however , in @xcite , the synchronization parameters are assumed to be constant and deterministic over the length of a block which is not a valid assumption for time varying phase noise .
a wiener filter approach that applies spatial correlation to improve phase noise estimation in mimo systems is proposed in @xcite .
however , the proposed solution is only applicable to uncoded mimo systems and the algorithm in @xcite introduces significant overhead to phase noise estimation process since it requires frequent transmission of _ orthogonal _ pilot symbols .
the problem of joint data detection and phase noise estimation for coded mimo systems over block fading channels is still unaddressed and will be the main focus of this thesis . in chapter 2 the phase noise model
is introduced and digital communication system for siso systems over the additive white gaussian noise ( awgn ) channel affected by phase noise is presented . in chapter 3 the performance of both uncoded and coded siso systems affected by phase noise
are investigated .
the iterative code - aided em - based approach used in @xcite is modified and derived analytically .
the em - based algorithm is implemented and its components are explained in detail .
two estimators that are proposed in @xcite , the ks - mla and the eks , are evaluated and their performances are compared against one another . in chapter 4 , the mimo system model for both uncoded and the coded mimo systems over rician fading channels in the presence of phase noise is described in detail .
an iterative joint phase noise estimation and data detection algorithm based on the em framework is derived analytically . a low complexity _ extended kalman filter - smoother _ ( ekfs ) is proposed to estimate the time varying phase noise processes of each oscillator .
bicm scheme is used to decrease the detection complexity .
the performance of the proposed algorithm is investigated via computer simulations . in chapter 5 conclusion and future research directions
are discussed .
the primary contributions of this thesis are summarized as follows : * the system model for both uncoded siso and uncoded mimo systems in the presence of phase noise are outlined in detail and an extended kalman filter with symbol - by - symbol feedback is proposed for each system .
the error rate performance of the proposed estimators are investigated through numerical results . *
the iterative code - aided em - based algorithm proposed in @xcite for coded siso systems is modified and derived analytically . moreover ,
the performances of two estimators , the ks - mla and the eks , are numerically compared with the help of computer simulations .
* an em - based receiver is proposed to perform iterative joint phase noise estimation and data detection for bicm - mimo systems . *
it is analytically demonstrated that a computationally efficient ekfs can be applied to carry out the maximization step of the em algorithm .
* a new low complexity soft decision - directed ekfs for tracking phase noise over the length of a frame is proposed and the filtering and smoothing equations are derived . * extensive simulations
are carried out for different phase noise variances to show that the performance of a mimo system employing the proposed receiver structure is very close to the ideal case of perfect knowledge of phase noise .
simulation results demonstrate that error rate performances of a 2@xmath02 los - mimo system using the proposed em - based receiver is very close to that of the perfectly synchronized system for low - to - medium signal - to - noise ratios .
it is also shown that the _ mean square - error _ ( mse ) of the phase noise estimates improves with every em iteration .
in wireless communication systems , the baseband signal is multiplied by a high frequency sine wave to operate at a certain frequency band , called carrier frequency .
local oscillators produce the carrier frequency waveforms .
phase - locked loop ( pll ) @xcite is the major phase recovery block in a communication system .
pll calculates the phase difference between the input and the output signal .
the difference is then filtered by a low pass filter and applied to the voltage controlled oscillator ( vco ) . the controlled voltage on the vco changes the oscillator frequency to minimize the phase difference of the input and the output signal . however , the output of the vco circuit is a non - ideal sine wave due to some hardware limitations .
the power spectrum of the output signal is not strictly concentrated at the carrier frequency .
the instantaneous output of a oscillator is given by @xcite @xmath1 where @xmath2 denotes the carrier frequency , @xmath3 denotes the amplitude , @xmath4 is amplitude noise and @xmath5 is phase noise .
demir _ et . al .
_ show in @xcite that amplitude noise decays over time , since the system stabilizes itself .
the amplitude noise may thus be ignored and the normalized oscillator output signal can be written as @xmath6 the oscillator phase noise can be seen as a widening of the spectral peak of the oscillator . the frequency domain single - sideband phase noise power , @xmath7 is defined as the ratio of the noise power in a 1hz sideband at an offset @xmath8 hertz away from the carrier , @xmath9 , to the total signal power , @xmath10 .
since we have no absolute time reference , the phase disturbances accumulate over time and can be represented by @xmath11 where @xmath12 is a white gaussian process with a constant power spectral density ( psd ) .
then , the phase noise process can be modeled as a wiener process and the oscillator power spectrum is a lorentzian , given by @xcite @xmath13 where @xmath14 denotes the 3db bandwidth .
it is seen that the spectrum is characterized by a single parameter , @xmath14 .
the phase noise process is sampled every @xmath15 seconds , sampling time interval .
then , the discrete time phase noise process is defined as , @xmath16 and can be modeled as a random walk in accordance with [ eqn : contphasenoise ] , i.e. discrete - time wiener process @xcite @xmath17 in ( [ eqn : thatadelta1 ] ) , the innovation term , @xmath18 is a discrete zero - mean gaussian random variable with variance @xmath19 , denoted as @xmath20 .
the phase noise innovation variance is given by @xcite @xmath21 note that the discrete innovation process is also white , @xmath22 in fig .
[ fig : figwienerpn ] a realization of the discrete time wiener phase noise process is plotted .
.,width=377 ] at the transmitter , a group of data bits are modulated onto an @xmath23-point quadrature amplitude modulation ( @xmath23-qam ) constellation @xmath24 , displayed in fig .
[ fig : fig16qam ] .
symbols are then transmitted through an awgn channel . in a communication system without the phase noise disturbances the received signal at time @xmath25 is given by @xmath26 where @xmath27 is the received signal , @xmath28 is the complex transmitted symbol , @xmath29 is the zero - mean awgn with variance @xmath30 per dimension , i.e. @xmath31 , as shown in fig .
[ fig : fig16qamawgn ] . the received signal is also effected by time varying phase noise both at transmitter and receiver .
let @xmath32}(k)$ ] and @xmath33}(k)$ ] denote the discrete time phase noise sample at transmitter and receiver , respectively .
the received signal at time @xmath25 is given by @xmath34}(k)}+\tilde w(k))e^{j\theta^{[r]}(k)}\\ & = & s(k ) e^{j(\theta^{[t]}(k)+\theta^{[r]}(k))}+ w(k)\label{eqn : recsign2}\\ & = & s(k ) e^{j\theta(k)}+w(k)\label{eqn : obspn}\end{aligned}\ ] ] where @xmath35}(k)}$ ] is the rotated noise sample , and @xmath36 is the total phase noise process .
note that rotation on the circular symmetric additive noise does not change the statistical properties , i.e. , @xmath37 .
the innovation of total phase noise process will have a zero - mean gaussian distribution and its variance will be the sum of the variances , @xmath38}}+\sigma^2_{\delta^{[r]}}\right)$ ] where @xmath39}}$ ] and @xmath40}}$ ] denote the innovation variance of the phase noise process at the transmitter an at the receiver , respectively . the total phase noise process rotates the signal constellation as displayed in fig [ fig : fig16qampn ] .
the received signal which is affected by the awgn and rotated by the phase noise is shown in fig .
[ fig : fig16qamawgnandpn ] .
the main focus of this thesis will be on the phase noise estimation for coded mimo systems . in this chapter , to better understand the effect of phase noise , the problem of phase noise estimation for siso systems over awgn channel is investigated .
it is assumed that perfect frame synchronization and phase recovery are performed at the beginning of each frame by transmitting sufficient number of pilot symbols .
therefore , the problem under consideration for siso systems is simplified to the problem of phase noise estimation before investigating the problem for coded mimo systems .
first , an uncoded siso system is taken under consideration in sec .
[ sec : uncodedsiso ] .
an extended kalman filter ( ekf ) is suggested to track time varying phase noise and the set of equations for the ekf is derived . in sec .
[ sec : codedsiso ] , the problem of joint phase noise estimation and detection for a coded siso system is discussed .
an algorithm is analytically derived from the em framework and is applied to iteratively solve the problem .
the enhancement of the ldpc codes and the kalman filters into the em - based algorithm is explained in detail . in order to de -
rotate the signal space and to achieve synchronous communication , time varying phase noise process should be estimated .
since the parameter to be estimated is not deterministic , an estimator based on the bayesian approach should be used @xcite . in the bayesian approach , prior knowledge about the random parameter
is also taken into account .
this approach is commonly used for the systems which can be represented with a dynamical model .
kalman filtering can be considered as a sequential minimum mean square error ( mmse ) estimator which works according to the bayesian framework .
the received signal in ( [ eqn : obspn ] ) and the phase noise process are used to construct a state space signal model .
the state and observation equations at time @xmath25 are given as @xmath41 where @xmath27 is the received signal , @xmath28 is the complex transmitted symbol belonging to the @xmath23-qam constellation , @xmath42 is the phase noise value , @xmath43 is the additive complex gaussian noise with variance @xmath44 and @xmath18 is the phase noise innovation , which is assumed to be gaussian distributed with variance @xmath19 . since the observation equation is nonlinear , a hard decision - directed ekf
is used instead @xcite .
the observation equation can be rewritten as @xmath45 where the nonlinear function @xmath46 is defined as @xmath47 in ( [ eqn : zfunc1 ] ) , @xmath48 is the hard decision of the transmitted symbol at time instance @xmath25 . note that for uncoded siso systems hard decision symbol , @xmath48 , is obtained by the demodulator at each time instance .
next , @xmath48 is input to the hard decision - directed ekf at each time instance . in other words , decision feedback is performed symbol - by - symbol .
the ekf provides phase noise estimate , @xmath42 .
the ekf first predicts the mean and the minimum prediction mse of the state ahead , @xmath49 and @xmath50 , respectively , given the previous values .
then , the ekf updates the estimates with the observation and computes the mean and the minimum mse of a posteriori state estimate , @xmath51 and @xmath52 , respectively .
the kalman gain , @xmath53 , indicates the amount of correction required for an observation sample .
since @xmath54 is a nonlinear function , it is linearized with a first - order taylor expansion .
therefore , @xmath55 denotes the jacobian of @xmath54 with respect to @xmath56 .
the first and the second moments of the state ahead are predicted as @xmath57 after the observation , the posteriori state estimate statistics are updated as @xmath58 where the kalman gain is determined as @xmath59 the jacobian of @xmath60 with respect to @xmath56 is given by @xmath61 in ( [ eqn : kalmangain ] ) , @xmath62 is the observation noise variance , given by @xmath63 the ekf needs to be properly initialized to compute @xmath64 and the mse @xmath65 .
it is assumed that a sufficient number of pilots is used for frequency , frame and phase synchronization at the beginning of the frame .
therefore , there is no phase shift of the first received signal i.e. @xmath66 .
moreover , at the first time instant , the minimum prediction mse is set to @xmath67 since this error amounts to the estimation of @xmath68 without any data .
the ekf is applied for different phase noise innovation variance values .
the bit error rate ( ber ) performance of the system is investigated .
ber vs. @xmath69 for 16-qam is shown in fig .
[ fig : fig1 ] .
it can be seen that as @xmath69 increases , the ekf performs better and can track phase noise processes with higher variances .
for 16-qam for various values of @xmath19.,width=453 ] in coded siso systems , the data bits are first encoded with an ldpc encoder .
then , the encoded bits are modulated onto complex symbols from an @xmath23-qam constellation @xmath70 and transmitted through the awgn channel .
the phase noise process should be estimated in order to de - rotate the signal space .
channel coding adds redundant bits to protect information bits and to correct erroneous bits .
a rate @xmath71 regular ldpc encoder from the nasa goddard technical standard @xcite is used .
it is a regular code with variable node degree 4 and check node degree 32 .
the ber vs. @xmath69 for both uncoded and coded systems without phase noise is shown in fig .
[ fig : nopncomparison ] .
data bits are mapped to 16-qam symbols . as expected ,
very low bers can be achieved with channel coding at medium and high @xmath69 levels .
for 16-qam uncoded and coded system without phase noise , width=415 ] in a coded siso system in the presence of phase noise , decoding and phase noise estimation can be performed jointly in an iterative way to complement one another .
the em framework is used in @xcite to estimate the constant phase offset . in @xcite , a modified em - based algorithm
is suggested to estimate time varying phase noise .
the tracking is performed by both an extended kalman filter and smoother ( eks ) and a linearized kalman filter and smoother with maximum - likelihood average ( ks - mla ) .
the mla algorithm removes the phase noise average over the frame . in @xcite ,
the modified em - based algorithm is not derived analytically . in the following ,
we demonstrate analytically how an em - based algorithm can be utilized for code - aided synchronization .
the performance of the proposed estimators is investigated numerically .
first , recall that the signal model and the phase noise model for the uncoded siso system is given by @xmath72 let @xmath73 denote the number of symbols in one frame .
we define the vectors @xmath74^t \\
\mathbf s & \triangleq & [ s(1 ) , s(2 ) , \dots , s(l_f)]^t \\
\theta & \triangleq & [ \theta(1 ) , \theta(2 ) , \dots , \theta(l_f)]^t .\end{aligned}\ ] ] given the observation vector , @xmath75 , _ the maximum a posteriori _
( map ) estimate of @xmath76 is given by @xmath77 where @xmath78 is the a priori probability of parameters @xmath76 .
the em framework is an iterative approach to solve an estimation problem where the observation , @xmath75 , depends not only on the parameter to be estimated , @xmath76 , but also on some nuisance parameter , @xmath79 .
note that transmitted symbols , @xmath79 , and phase noise @xmath76 are independent .
the em algorithm iteratively maximizes the conditional a posteriori expectation , @xmath80 , where @xmath81 is the conditional log likelihood function ( llf ) and @xmath82 denotes the expectation operator . in the expectation
step the block of previous phase noise estimates , @xmath83 , is kept fixed and it is updated in the maximization step .
the em framework for the @xmath84th iteration is defined by @xcite @xmath85 the em algorithm has been shown to converge to the map solution if it is accurately initialized . given the transmitted data and the phase noise process , the received signal has a gaussian distribution .
after taking logarithm and dropping constant terms , the conditional llf is given by @xmath86 taking the conditional expectation over @xmath79 , given @xmath75 and @xmath83 , equation for the expectation step can be written as @xmath87 where the weighted or soft symbol @xmath88 is defined as @xmath89 in ( [ eqn : alpha1]),@xmath90 are the marginal a posteriori symbol probabilities ( apps ) of the transmitted symbol . for 16-qam ,
if gray mapping is used , ( [ eqn : alpha1 ] ) can be computed as @xmath91 where @xmath92 and @xmath93 is the log likelihood ratio ( llr ) of the a posteriori probability of the @xmath84 th bit of the symbol at time instant @xmath25 , @xmath94 .
the llr is defined as @xmath95 the block diagram of the receiver structure is shown in fig .
[ fig : figrs_siso ] where superscript @xmath96 denotes the iteration of the em algorithm .
the llrs are first computed by a soft demapper ( demodulator ) .
then , they are passed to the ldpc decoder .
next , the ldpc decoder runs a sufficient number of iterations to achieve more accurate llrs .
after that , the llrs computed by the ldpc decoder are modulated into soft symbols according to ( [ eqn : alpha ] ) by a soft mapper . as a result
, the expectation step computes the block of apps denoted as @xmath97^t .\end{aligned}\ ] ] the maximization step maximizes @xmath98 in ( [ eqn : mstepsiso ] ) with respect to @xmath76 . since the system can be modeled in a state - space form , the solution can be given by kalman filtering .
the kalman filter is an optimal mmse estimator which can be considered as a map estimator @xcite .
then , a decision directed kalman filter where the transmitted symbol vector , @xmath79 , is replaced by the soft decision symbol vector , @xmath99 , estimates @xmath76 as @xmath100 by setting @xmath101 , the conditional llf in ( [ eqn : lnestepsiso ] ) can be rewritten as @xmath102 using ( [ eqn : lnestepsiso2 ] ) in ( [ eqn : mapthetasiso ] ) we achieve ( [ eqn : qfunc ] ) which is the function to be maximized in the m - step . as a result ,
a decision directed kalman filter can be used to carry the m - step . in @xcite ,
two soft decision - directed kalman filters , the ks - mla and the eks , are proposed . in the following , two estimators will be explained .
[ [ the - kalman - filter - and - smoother - with - maximum - likelihood - average ] ] the kalman filter and smoother with maximum likelihood average + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the ks - mla proposes a new observation equation which is a linear function of @xmath56 . assuming that @xmath103 , the ks - mla first removes the data dependency from the measurement via @xmath104 where @xmath105 is the new observation equation .
then , the ks - mla assumes that the phase noise values are small enough such that @xmath106 subsequently , new observation equation is given by @xmath107 where @xmath108 and @xmath109 is a zero mean gaussian random variable with variance @xmath110 and @xmath111 is the average symbol energy . since the observation equation is linear , a regular kalman filter and smoother ( ks ) can be used .
note that the new observation equation does not depend on the complex transmitted symbol .
instead , it is a function of the amplitude square of the soft symbol .
phase noise process reaches large values within a block .
depending on @xmath19 and @xmath73 , it has a nonzero average over the block .
the mla algorithm estimates the ml average of the phase noise process over each block and removes it from each received block , @xmath75 , at the input of ks .
then , the parameter to be estimated is given by @xmath112 where @xmath113 the mla algorithm is directly embedded into the em framework and performed before the estimator in the m - step .
[ [ extended - kalman - filter - and - smoother ] ] extended kalman filter and smoother + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + second estimator proposed in @xcite is the soft decision - directed eks .
the observation equation in ( [ eqn : obsuncoded ] ) is a nonlinear function of @xmath56 .
therefore , a suboptimal decision directed eks is used to track the phase noise process in @xcite .
the state and observation equations at time @xmath25 are given as @xmath114 note that ( [ eqn : obsuncoded ] ) and ( [ eqn : obscoded ] ) are similar .
we can rewrite the observation equation as @xmath115 where @xmath116 stands for the nonlinear function for the coded system .
the eks approximates @xmath116 as @xmath117 where @xmath118 is the estimate of @xmath42 based on the previous data .
the eks first filters the received signal and estimates the block of parameters .
afterwards , it smooths the estimations with a backward recursion .
it is initialized with state estimate @xmath66 and a posteriori mse @xmath67 .
the filtering equations compute @xmath119 and @xmath120 in a recursive fashion according to ( [ eqn : hattheta1]-[eqn : ck ] ) where @xmath54 is replaced by @xmath121 .
the smoother runs a backward recursion to find better estimates of the parameter block .
we will use subscript @xmath122 to refer to smoothing .
@xmath123 denotes the sequence from @xmath25 to @xmath73 .
the set of equations for smoothing are given by @xmath124 ldpc codes are linear block codes that have a sparse parity check matrix @xmath125 .
the input of the ldpc decoder are the llrs that are computed by the soft demapper ( demodulator ) .
the output of the ldpc decoder are the updated llrs that will be used by the soft mapper ( modulator ) .
an example of a parity check matrix of a ( 7,4 ) irregular block code , i.e. the number of 1 s in each row and in each column is not constant , is given by @xmath126.\label{eqn : hldpc}\end{aligned}\ ] ] the corresponding tanner graph is shown in fig .
[ fig : tanner1 ] .
there are 7 variable nodes and 4 check nodes for @xmath125 .
let @xmath127 an @xmath128 be the index of the rows and columns of @xmath125 , respectively . if @xmath125[k , l ] is 1 , then there exists a connection between the check node @xmath25 and the variable node @xmath129 .
note that the information exchange is done in two ways for each connection , one from variable node to check node and one for the check node to variable node .
in ( [ eqn : hldpc]).,width=415 ] the ldpc decoder runs the sum product algorithm to estimate the llrs of the transmitted bits .
it computes the llr of each path from variable nodes to check nodes , and from check nodes to variable nodes iteratively .
@xmath130 denotes the llr of the @xmath129th bit of the transmitted codeword @xmath131 .
let @xmath132 denote the llr that belongs to the connection in the direction of check node @xmath25 to variable node @xmath129 .
similarly , @xmath133 denotes the llr of the same connection in the opposite direction .
an example is also shown in fig . [ fig : tanner1 ] .
@xmath134 denotes the updated llr of the @xmath129th bit of the transmitted codeword .
the sum product algorithm is performed as follows @xmath135 where @xmath136 represents the set of indexes of all the variable nodes connected to the check node @xmath25 except for the variable node @xmath129 .
similarly , @xmath137 represents the set of indexes of all the check nodes connected to the variable node @xmath129 except for the check node @xmath25 . for example , if @xmath138 and @xmath139 , @xmath140 , @xmath141 and @xmath142 .
[ [ improving - the - speed - of - the - algorithm ] ] improving the speed of the algorithm + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + @xmath133 should be initialized for the first iteration .
normally , it is initialized with the input llrs as @xmath143 as suggested in @xcite , instead of initializing with the input llrs , @xmath133 is initialized by the last update of it from the latest em algorithm iteration .
thus , @xmath144 where @xmath96 denotes the algorithm iteration .
it is seen that significant complexity reduction can be achieved with negligible degradation in ber with this approximate version .
the number of the decoder iterations can be reduced and the number of the em algorithm iterations can be increased .
since the complexity heavily depends on the number of decoder iterations , the overall complexity is much less than that of the original em - based algorithm .
first , the advantage of keeping internal information at the decoder is investigated for the 256-qam system with the eks . in fig .
[ fig : decode_no_restart_256 ] , the ber is calculated for each em iteration for different number of decoding iterations with and without keeping internal information at the decoder for a fixed @xmath145db .
the dashed lines represent the algorithm where the ldpc decoder is reinitialized between each iteration of the em algorithm . in fig .
[ fig : decode_no_restart_256 ] , the solid lines represent the new algorithm where the internal information in the ldpc decoder is kept between the em algorithm iterations .
it is seen that with keeping internal information overall complexity can be reduced due to fewer iterations .
when the ldpc decoder is reinitialized the ber tends to reach a floor and the system can not perform under this level even with more em iterations . on the other hand ,
when we keep the internal information at the ldpc decoder , lower bers can be achieved with increasing number of em iterations since better llrs are obtained from the decoder in each em iteration .
db for different number of decoder iterations , with and without keeping internal decoder information.,width=415 ] in fig .
[ fig : bervssnr1e43dec ] , the results of the em algorithm are shown for different number of iterations for both the eks and the ks - mla for a 16-qam system where @xmath146 . the number of iterations performed by the ldpc decoder is set to 3 .
it is seen that the ks - mla performs slightly better than the eks .
for 16-qam coded system with both the eks and the ks - mla where @xmath146 and 3 decoding iteration.,width=415 ] the ks - mla suggests a new linear observation equation ( [ eqn : newobs ] ) by assuming @xmath147 .
moreover , the received signal is multiplied by the soft decision symbol which also increases the noise power .
ad hoc _ method is proposed in @xcite to alleviate the effects of the assumptions in ( [ eqn : sintheta ] ) and ( [ eqn : costheta ] ) .
when the phase noise values are small , @xmath88 is a reliable estimate of @xmath28 .
then , the ks - mla algorithm performs better than the eks
. however , the ks - mla fails to track more severe phase noise processes . in fig .
[ fig : bervssnr3e43dec ] , the phase noise innovation variance is set to @xmath148 where all the other parameters remain unchanged . it is concluded that the assumptions ( [ eqn : sintheta ] ) and ( [ eqn : costheta ] ) are violated and @xmath88 is a less reliable estimate of @xmath28 . thus , the kfs - mla performs significantly worse than the eks .
moreover , the ks - mla has an irreducible error floor . for 16-qam coded system with both the eks and the ks - mla where @xmath148 and 3 decoding iteration.,width=415 ] in fig .
[ fig : bervssnr1e43dec_256 ] , the results for both the eks and the ks - mla are shown for a 256-qam system where @xmath146 . it is well - known that the reliability of the estimate of a 256-qam symbol is less than that of a 16-qam symbol for the same phase noise process .
the reason is that phase noise effects the outer symbols in the signal constellation .
in addition , the 256-qam constellation is more densely packed than the 16-qam constellation , i.e. , the distance and the phase difference between two neighbor constellation points is smaller . therefore , the performance of the ks - mla is degraded since the noise power is also amplified . in fig .
[ fig : bervssnr1e43dec_256 ] , we observe that the eks performs better than the ks - mla . as a result ,
the eks is more advantageous than the ks - mla for higher order modulations .
for 256-qam coded system with both the eks and the ks - mla where @xmath149 and 3 decoding iteration.,width=415 ]
the demand for high data rate wireless communications has spearheaded the research in the field of mimo systems .
it is shown that the bandwidth efficiency of a wireless link can be significantly improved with the usage of mimo systems @xcite .
however , synchronization errors decrease the performance of mimo system dramatically . in sec .
[ sec : uncodedmimo ] the system model for an @xmath150 uncoded mimo systems is introduced .
detection in mimo systems is also briefly discussed .
an ekf is applied to track multiple phase noise processes . in sec .
[ sec : codedmimo ] , the main contribution of this thesis is provided .
joint phase noise estimation and detection in coded mimo systems is performed with the help of an em - based algorithm for the first time in the literature .
exploiting ldpc codes , ekfs and bicm , the proposed em - based algorithm iteratively solves the problem of joint phase noise estimation and detection .
simulation results are also evaluated for a @xmath151 los - mimo system over rician fading mimo channels .
an uncoded mimo system with @xmath152 transmit and @xmath153 receive antennas is under consideration . at the transmitter ,
a group of data bits are modulated onto an @xmath23-qam constellation @xmath24 .
they are demultiplexed into substreams of symbols of length @xmath73 . subsequently , using spatial multiplexing the symbols are transmitted simultaneously from @xmath152 antennas .
quasi - static block fading channels are considered , i.e. , the channel gains remain constant over the length of a frame but change from frame to frame .
similar to previous work in the literature @xcite , it is assumed that the @xmath154 channel matrix @xmath155 is estimated using orthogonal training sequences that are transmitted at the beginning of each frame . to ensure that the proposed scheme is applicable to los and multi - user mimo systems
, it is assumed that independent oscillators are deployed at each transmit and receive antenna .
awgn is also taken into consideration . the received signal is also effected by time varying phase noise both at the transmitter and the receiver side . to ensure that the proposed scheme is applicable to los and multi - user mimo systems
, it is assumed that independent oscillators are deployed at each transmit and receive antenna .
the performance of the uncoded mimo system is severely degraded in the presence of phase noise . in @xcite , pilot - aided estimation of phase noise in an uncoded mimo system is investigated .
the phase noise parameters corresponding to the transmit and receive antennas are estimated by applying a wiener filtering approach . however , the proposed scheme is bandwidth inefficient and significant overhead is introduced . in @xcite ,
joint channel and phase noise estimation is performed . a data aided least square ( ls ) estimator , a decision directed weighted least squares ( wls ) estimator and a decision directed ekf are proposed to track time varying phase noise . based on the above set of assumptions , the received signal at time instance @xmath25 , @xmath156^t$ ] , is given by @xmath157}(k ) \mathbf h \boldsymbol \gamma^{[t]}(k ) \mathbf s(k ) + \mathbf w(k)\end{aligned}\ ] ] where * @xmath158}(k)\triangleq \textrm{diag}\big(e^{j\theta_1^{[r]}(k)},e^{j\theta_2^{[r]}(k)},\dots , e^{j\theta_{n_r}^{[r]}(k ) } \big)$ ] is an @xmath159 diagonal matrix and @xmath160}(k ) \triangleq \textrm{diag}\big(e^{j\theta_1^{[t]}(k)},e^{j\theta_2^{[t]}(k)},\dots , e^{j\theta_{n_t}^{[t]}(k ) } \big)$ ] is an @xmath161 diagonal matrix , * @xmath162}(k)$ ] and @xmath163}(k)$ ] denote the phase noise process at the @xmath164th receive and @xmath165th transmit antennas , respectively , where the @xmath154 mimo channel matrix is denoted as @xmath166 $ ] with @xmath167^t$ ] * @xmath168 , for @xmath169 and @xmath170 , denotes the channel gain between transmit antenna @xmath165 and receive antenna @xmath164 that is modeled as a complex gaussian random variable , @xmath171 * @xmath172^t$ ] is the vector of transmitted symbols , * @xmath173^t$ ] is the vector of zero - mean additive white gaussian noise ( awgn ) , @xmath174 .
phase noise is modeled as discrete - time wiener process . therefore , for @xmath169 and @xmath170 @xcite @xmath175}(k)=&\theta_\ell^{[r]}(k-1)+\delta_\ell^{[r]}(k)\notag \\
\theta_m^{[t]}(k)=&\theta_m^{[t]}(k-1)+\delta_m^{[t]}(k),\end{aligned}\ ] ] where @xmath176}(k)$ ] and @xmath177}(k)$ ] are the phase innovations for the @xmath164th receiver and @xmath165th transmit antenna , respectively .
they are assumed to be white real gaussian processes with @xmath176}(k)\thicksim \mathcal n(0,\sigma^2_{\delta_\ell^{[r]}})$ ] and @xmath177}(k)\thicksim \mathcal n(0,\sigma^2_{\delta_m^{[t]}})$ ] .
let us define @xmath178 as @xmath179}(k ) , \dots , \theta_{n_r}^{[r]}(k ) , \theta_1^{[t]}(k ) , \dots , \theta_{n_t}^{[t]}(k ) \right]^t.\end{aligned}\ ] ] then we have @xmath180 where @xmath181}(k ) , \dots , \delta_{n_r}^{[r]}(k ) , \delta_1^{[t]}(k ) , \dots , \delta_{n_t}^{[t]}(k)\right]^t$ ] .
we want to detect the transmitted symbol vector @xmath182 in the maximum - likelihood ( ml ) sense . for an @xmath23-qam modulated @xmath150 mimo system , ml detection of the transmitted symbol vector at time instant @xmath25
is given by @xmath183 where @xmath184}(k ) \mathbf h \boldsymbol \gamma^{[t]}(k ) \mathbf s.\end{aligned}\ ] ] the ml detection search over @xmath185 combinations , @xmath186 , of the transmitted symbol vector .
the problem in ( [ eqn : uncmimodetection ] ) is computationally infeasible especially for higher order modulations .
there are several detectors that provide approximate solutions with low complexity , such as zero - forcing detection with decision feedback @xcite , sphere detection @xcite , and lattice reduction aided detection @xcite
. however , low complexity mimo detection algorithms are beyond the scope of this thesis . the ml detection problem in ( [ eqn : uncmimodetection ] ) is directly solved by searching over all possible transmitted symbol vectors .
the ml detector provides hard decision symbol vector , @xmath187 .
next , @xmath187 is input to the ekf which estimates phase noise vector , @xmath178 . a hard decision directed ekf
is applied to estimate phase noise parameters .
let us define @xmath188}(k ) \mathbf h \boldsymbol \gamma^{[t]}(k)$ ] .
consequently , the vector of received signals at the receive antennas at time @xmath25 , @xmath189 can be rewritten as @xmath190 it can be seen from ( [ eqn : statemimo1 ] ) and ( [ eq : obsmimounc2 ] ) that there are a total of @xmath191 phase noise processes that need to be tracked .
however , as shown below , an equivalent signal model can be obtained that reduces the dimensionality of the estimation problem and results in reduces overhead and improved estimation accuracy @xcite . by arbitrarily selecting the phase noise process @xmath192}(k)$ ] as a reference phase noise value , @xmath193 in ( [ eq : received_signal2 ] )
can be rewritten as @xmath194}(k)\right)}{\boldsymbol \gamma}^{[r]}(k ) \mathbf h { \boldsymbol \gamma}^{[t]}(k)e^{-j\left(\theta_{n_t}^{[t]}(k)\right ) } \notag\\ & = \tilde { \boldsymbol \gamma}^{[r]}(k ) \mathbf h \tilde { \boldsymbol \gamma}^{[t]}(k),\end{aligned}\ ] ] where * @xmath195}(k)\triangleq \textrm{diag}\big\{e^{j(\phi_1(k ) ) } , \dots , e^{j(\phi_{n_r}(k))}$ ] } , * @xmath196}(k)\triangleq \textrm{diag}\big\{e^{j(\phi_{n_r+1}(k ) ) } , \dots , e^{j(\phi_{n_r+n_t-1}(k ) ) } , 1 \big\}$ ] , * @xmath197^t$ ] , and @xmath198}_f+\theta^{[t]}_{n_t}&f=1,\cdots , n_r,\\ \theta^{[t]}_{f - n_r}-\theta^{[t]}_{n_t}&f = n_r+1,\cdots , n_r+n_t-1 .
\end{cases } \end{aligned}\ ] ] it is seen that @xmath191 state parameters can be replaced to @xmath199 new state parameters without changing the observation equalities . then , we can rewrite the state and observation equations at time @xmath25 as @xmath200 where @xmath201^t,\nonumber\\ & = & \tilde { \boldsymbol \gamma}^{[r]}(k ) \mathbf h \tilde { \boldsymbol \gamma}^{[t]}(k ) \hat { \mathbf s}(k),\nonumber\end{aligned}\ ] ] and @xmath202 , denotes the @xmath165th element of vector @xmath203 .
note that we assume that @xmath204 is known , @xmath187 is provided by the detector and @xmath205 is a nonlinear function of @xmath206 .
it is assumed that perfect phase synchronization is obtained at the beginning of a frame by transmitting sufficient number of pilots .
then , the ekf is initialized with the state estimate @xmath207 and the error covariance estimate @xmath208 .
the ekf first estimates the a priori state vector @xmath209 , and the @xmath210 a priori error covariance matrix @xmath211 as follows @xmath212 next , the ekf linearizes the nonlinear function @xmath203 in ( [ eqn : obsmimounc42 ] ) about the a priori estimate of the state vector using a first - order taylor approximation as @xmath213 where the @xmath214 jacobian matrix with respect to @xmath215 at time instance @xmath25 , @xmath216 , is defined as @xmath217 the jacobian matrix can be constructed with two matrices as @xmath218\end{aligned}\ ] ] where the @xmath159 matrix @xmath219 is defined as @xmath220 and the @xmath221 matrix @xmath222 is given by @xmath223.\end{aligned}\ ] ] after the observation , the a posteriori estimate of the state vector , @xmath224 , and the error covariance matrix , @xmath225 , are given by @xmath226 where the @xmath227 kalman gain matrix , @xmath228 , is given by @xmath229 finally , the observation noise covariance matrix is given by @xmath230 for 2x2 uncoded mimo system with bpsk modulation for various @xmath231 values.,width=453 ] ber vs. @xmath69 for bpsk modulation and ml equalization for a 2x2 mimo system where @xmath232 is shown in fig .
[ fig : fig2mimo ] .
it is seen that the ekf performs close to the synchronized system at low to medium @xmath69 levels where @xmath233 . on the other hand
, the ber performance of the system degrades at high @xmath69 levels .
note that the number of parameters to be tracked by the ekf is larger than the number of observation equations .
this results in an error floor , i.e. the performance of the system can not be improved with increasing @xmath69 .
the error floor depends on the phase noise innovation variance , @xmath231 . in fig .
[ fig : fig2mimo ] we observe that error floor tends to occur at lower signal - to - ratios for the system with @xmath234 .
mimo systems can utilize the available spectrum more efficiently @xcite,@xcite .
a number of coding structures have been proposed in order to achieve high spectral efficiency , including multi - dimensional trellis - coded modulation @xcite , space time block codes @xcite , mimo multilevel coding @xcite .
bicm is also one of the popular scheme that enables communication systems to fully exploit the spectrum efficiently @xcite .
code - aided synchronization based on the em framework for joint channel estimation , frequency and time synchronization for a bicm - mimo system is proposed in @xcite . however , in @xcite , the synchronization parameters are assumed to be constant and deterministic over the length of a block , which is not a valid assumption for time varying phase noise . in the following , we show analytically how an em - based algorithm with a soft decision - directed ekfs can be used to estimate and compensate time varying phase noise .
first , we recall the signal model for uncoded mimo systems .
the received signal at time instance @xmath25 is given by , @xmath235}(k ) \mathbf h \boldsymbol \gamma^{[t]}(k ) \mathbf s(k ) + \mathbf w(k)\nonumber\\ & = & \mathbf x(k ) { \mathbf s}(k ) + \mathbf w(k)\nonumber\end{aligned}\ ] ] and the new state vector of length @xmath236 can be written as @xmath237 the em algorithm iteratively solves an estimation problem where the observation , @xmath238_{n_r \times l_f } \triangleq [ \mathbf y(1 ) , \mathbf y(2 ) , \dots , \mathbf y(l_f)]$ ] , depends not only on the parameters to be estimated , @xmath239_{(n_r+n_t ) \times l_f } \triangleq [ \boldsymbol \theta(1 ) , \boldsymbol \theta(2 ) , \dots , \boldsymbol \theta(l_f)]$ ] , but also on some unknown parameters , @xmath240_{n_t \times l_f } \triangleq [ \mathbf s(1 ) , \mathbf s(2 ) , \dots , \mathbf s(l_f)]$ ] @xcite .
given the observation sequence @xmath241 , the map estimate of @xmath242 is given by @xcite @xmath243 where @xmath244 .
the em algorithm consists of the _ expectation step _ ( e - step ) and _ maximization step _ ( m - step ) @xcite . for the @xmath84th em iteration ,
the e - step and m - step equations are given by @xmath245 respectively .
the em algorithm converges to the map solution if the initial estimates of the parameters of interest , @xmath246 , are sufficiently close to the true values of the parameters .
otherwise , the em algorithm may converge to a saddle point or a local maximum . to ensure the convergence ,
pilot symbols are inserted into the data stream every @xmath247 time instances . in the following subsections
we derive the e - step and m - step of the em algorithm for coded mimo systems .
let us define @xmath188}(k ) \mathbf h \boldsymbol \gamma^{[t]}(k)$ ] .
consequently , the vector of received signals at the receive antennas at time @xmath25 , @xmath189 can be rewritten as @xmath248 according to , the llf of the received signal matrix , @xmath249 , given the transmitted data , @xmath250 , and the phase noise process , @xmath251 , is given by @xmath252 using in the e - step of the em algorithm can be determined as @xmath253 where * @xmath254 and * @xmath255 . in , @xmath256 denotes the marginal posterior mean of the coded symbol vector at time instance @xmath25 , @xmath257 denotes the apps of the coded symbol vector given @xmath241 and @xmath258 , and @xmath259 is an ( @xmath161 ) matrix . as a result , the e - step reduces to the computation of the apps .
note that the apps can be computed via an iterative map decoder which is outlined in section [ sec : apps ] . as shown in
, the m - step maximizes the e - step with respect to the parameter of interest , @xmath260 . in the following
, it is shown that a low complexity soft decision - directed ekfs can be used to carry out the m - step of the em algorithm .
note that a kalman filter is equivalent to a map estimator @xcite . as a result , the kalman filter estimates of @xmath260
, @xmath261 can be written as @xmath262 where @xmath263 $ ] is an @xmath264 matrix . by setting @xmath265 , ( [ eqn : channellf ] )
can be rewritten as @xmath266 note that ( [ eqn : channellf2 ] ) and the e - step equation in ( [ eqn : estepfinal ] ) are equivalent to one another except for the term @xmath267 . however , when the soft decisions reach their true values , we can assume that @xmath268 .
thus , using and , we can conclude that a kalman filter can be applied to carry out the m - step of the em algorithm .
note that the _ observation _ equation for the kalman filter is given by while based on , the _ state _ equation at time @xmath25 is given by @xmath269 where @xmath181}(k ) , \dots , \delta_{n_r}^{[r]}(k ) , \delta_1^{[t]}(k ) , \dots , \delta_{n_t}^{[t]}(k)\right]^t$ ] .
the transmitted symbols @xmath182 , can be replaced by their a posteriori means , i.e. , soft decisions @xmath256 computed at the e - step using the iterative map detector in section [ sec : apps ] .
subsequently , the _ observation _ equation in can be rewritten as @xmath270 since the _ observations _ are a nonlinear function of the parameters of interest , @xmath271 , an _ extended _ kalman filter - smoother needs to be used instead to carry out the m - step of the em algorithm @xcite .
note that ( [ eq : obsmimounc2 ] ) and ( [ eqn : received_approxmc ] ) are similar
. we can rewrite the observation equation for the coded mimo system as @xmath272 the ekfs first filters the received signal and provides phase noise estimates with a forward recursion over the frame .
the filtering equations compute the posteriori estimate of the state vector , @xmath273 , and the error covariance matrix , @xmath274 in a recursive fashion according to ( [ eqn : hatphimimo]-[eqn : hatmpostmimo ] ) where @xmath203 is replaced by @xmath275 . in other words , instead of the hard decision vector provided by the ml detector in the uncoded mimo system , @xmath187 , the soft decision vector computed by the iterative detector , @xmath88 ,
is provided to the ekfs for the coded mimo system .
the ekfs is initialized with the state estimate @xmath207 and the error covariance estimate @xmath208 .
after filtering is completed for the whole block , a backward recursion is performed to smooth the estimates .
finally , the smoothed estimates of the phase noise parameters are carried out to be fed to the detector .
the smoothed estimate of the phase noise vector , @xmath224 , and the error covariance matrix @xmath225 are given by @xmath276 after the backward recursion is completed the block of phase noise estimates @xmath277 , is fed to the iterative detector for the next em algorithm iteration .
it is shown in sec .
[ sec : mstep ] that soft decisions , i.e. , the marginal posterior probabilities of the coded symbol vectors @xmath278 are required for the ekfs .
the computation of the true posterior probabilities has a complexity that increases exponentially with the frame length @xmath73 .
therefore , a near optimal iterative detector , operating according to the turbo principle @xcite,@xcite,@xcite is used to obtain the marginal a posteriori bit probabilities given the phase noise estimates and the channel gain matrix @xmath204 .
then , a soft modulator maps the a posteriori bit probabilities to symbol probabilities and constructs the soft decisions . the block diagram of the proposed em - based receiver structure , including both the ekfs and the iterative detector , is shown in fig .
[ fig : fig3 ] .
the iterative detector first computes @xmath185 conditional likelihoods of the symbol vectors given the phase noise estimates as @xmath279 note that the conditional likelihoods are computed by the equalizer from the received signal before the execution of the detector iterations .
iterative detection is then performed with 3 nested iterations , as seen in fig .
[ fig : fig3 ] . the iterative part of the detector operates according to the turbo principle .
the conditional _ a posteriori _ probability of the transmitted symbol at the transmit antenna @xmath165 , at time instance @xmath25 , @xmath280 , is factored as the product of an _ a priori _ probability ( with subscript @xmath281 ) , and an _ extrinsic _ probability ( with subscript @xmath282 ) such as @xmath283 where @xmath284 is a normalization constant .
note that since an interleaver at the transmitter side is used , the transmitted symbols on each antenna and the transmitted bits within each constellation symbol are independent . the extrinsic symbol probabilities in ( [ eqn : symbpostfact ] ) are computed by the equalizer in fig .
[ fig : fig3 ] .
the extrinsic symbol probability given the estimated phase noise parameters and the received signal @xmath285 for @xmath286 , is given by @xmath287 the a priori symbol probabilities @xmath288 are computed by the soft modem and fed to the equalizer after decoding is performed . therefore , the first iteration is between the equalizer and the soft modem and there are @xmath289 iterations .
the soft modem consists of the demapper , the interleaver , the iterative map decoder , the deinterleaver and the mapper as it is shown in fig .
[ fig : fig3 ] .
the demapper takes the extrinsic _ symbol _ probabilities from the equalizer and computes the extrinsic _ bit _ probabilities .
the bit posterior probabilities of the @xmath290th bit of the bit sequence mapped to the symbol @xmath280 , denoted as @xmath291 , is factored similar to ( [ eqn : symbpostfact ] ) as @xmath292 where @xmath293 is a normalization constant .
then , the extrinsic bit probability for @xmath294 is computed by the demapper as @xmath295 the demapper also needs the a priori _ bit _ probabilities computed by the map decoder .
the second iteration of the 3 nested iterations of the detector is performed between the demapper and the decoder @xmath296 times .
after the demapper computes all the extrinsic bit probabilities , they are deinterleaved and provided to the map decoder by the deinterleaver .
the decoder takes the deinterleaved extrinsic bit probabilities and computes both the deinterleaved _ posterior _ and the deinterlaved _ a priori _ bit probabilities exploiting code properties with @xmath297 iterations .
the posterior bit probabilities @xmath298 are sent out of the detector , interleaved and used to construct soft decisions for the ekfs , described in section [ sec : ekfs ] .
the posterior bit probabilities are also used for hard decision after the algorithm terminates .
the a priori bit probabilities @xmath299 are sent to the interleaver which constructs the a priori bit probabilities that are used as a priori information by the demapper .
finally , the mapper converts a priori bit probabilities to a priori symbol probabilities which are used as a priori information by the equalizer according to @xmath300 estimator.,width=340 ] aforementioned , the iterative detector provides the ekfs with the marginal posterior probabilities of the coded symbol vectors , @xmath301 .
the e - step is finished by the computation of these probabilities as @xmath302 where @xmath303 and @xmath304 are normalization constants .
then , these posterior probabilities are used to construct the soft decision , @xmath256 .
the a priori symbol probabilities @xmath305 and bit probabilities @xmath288 are initialized with a uniform distribution at the first em algorithm iteration
. a sufficient number of iterations inside of the detector is required for convergence if the detector is reinitialized with uniform probabilities at each em iteration . instead
, the detector can be initialized with the a priori probabilities obtained at the previous em iteration .
in addition , only 1 iteration is allowed inside of the iterative detector for each nested iteration , i.e. , @xmath306 as suggested in @xcite .
this alternative approach requires more em iterations but less detector iterations to converge . in this way
, the computational complexity of both the detector and the receiver can be reduced significantly . , and da initial estimation with @xmath307.,width=340 ] at the transmitter ,
data bits are first encoded by a rate @xmath71 regular ldpc encoder from the nasa goddard technical standard @xcite .
it is a regular code with variable node degree 4 and check node degree 32 .
the number of data bits in each frame , @xmath308 is equal to 7154 .
then , encoded bits are modulated onto 16-qam symbols .
therefore , there are @xmath309 symbol vectors in each frame .
performance will be measured as a function of @xmath69 , where @xmath310 denotes the transmitted energy per information bit and @xmath311 is the power spectral density of the awgn , i.e , @xmath312 .
the convergence of the em based estimator is severely dependent on the initialization of the estimation parameters .
first , the em based algorithm is initialized by a data aided ( da ) estimator to provide the initial phase noise estimates .
pilots are inserted into the data stream every @xmath247 time instance . at the first em algorithm iteration ,
the ekfs makes use of the pilots and estimates phase noise values at each @xmath247 time instance . afterwards
a linear interpolation is performed between two consecutive phase noise estimates . as a result ,
initial phase noise estimates are obtained and sent to the iterative detector to initialize the em - based algorithm .
da@xmath313 denotes the da estimator with pilot rate @xmath247 . for several phase noise processes and em algorithm iterations.,width=340 ] in fig .
[ fig : berpr14 ] , the effect of the phase noise on the ber performance corresponding to the em - based algorithm with da@xmath314 estimator for different phase noise innovation variance levels , @xmath19 , after the 3rd em algorithm iteration is shown .
as compared to perfect synchronization , no phase noise scenario , the proposed em - based algorithm gives rise to a ber degradation of about 1db in the presence of the slowly time varying phase noise , i.e. , @xmath315 .
the ber degradation amounts to about 2db when @xmath316 .
it is possible to track a stronger phase noise process with innovation variance @xmath317 in expense of 4db of @xmath69 comparing to perfectly synchronized system .
note that the performance of the system tends to stay constant as @xmath69 increases , yielding an error floor .
the main reason is that the prediction error of the ekfs is determined by the phase noise innovation variance . as @xmath19 increases , the error floor is observed at higher error rates . for several
number of decoder iterations , @xmath318.,width=340 ] code - aided synchronization techniques offer to improve the overall accuracy and performance of the system .
first , we investigate the performance of the em - based algorithm at each em algorithm iteration
. the em - based algorithm does not converge to the global solution at some of the erroneous frames resulting in an arbitrary large number of erroneous bits .
for this reason the ber performance of the system may not increase at each em algorithm iteration .
therefore , the fer performance of the system is investigated .
[ fig : fer2emits ] shows the fer performance at several em algorithm iterations for @xmath319 .
we observe that the fer performance of the system improves at each em algorithm iteration . in order to operate the system at @xmath320 fer at the 10th iteration , additional @xmath69 of about 2db
is required .
secondly , we investigate the estimation accuracy of the em - based algorithm .
the mse of the phase noise estimates at the first receive antenna is averaged over the successfully decoded frames .
[ fig : mseemits ] shows the mse results of different levels of phase noise innovation variance at both the 1st and the 5th em algorithm iteration .
it is seen that the proposed em - based algorithm not only increases the overall performance of the system but also yields better estimates at each em algorithm iteration .
we also investigate the performance of the proposed em - based algorithm for the system that is affected by a severe phase noise process such that @xmath318 .
[ fig : ferldec ] shows the effect of the number of decoder iterations on the fer performance of the system for a fixed @xmath321 .
we observe that the performance of the system does not improve significantly after a few em algorithm iterations .
in contrast , the fer performance of the system can be slightly improved with more iterations inside of the decoder . in order to achieve lower error rates in the presence of strong phase noise
a stronger channel encoder can be used , i.e , the rate of the ldpc code , @xmath322 , needs to be decreased .
note that the spectral efficiency of the system also reduces with decreasing @xmath322 yielding low throughput .
, da initial estimation with @xmath307 , and @xmath323 rate code.,width=340 ] fig [ fig : fer5e4rate12 ] shows the fer performance of the em - based algorithm for @xmath318 where data bits are encoded by a rate @xmath323 ldpc encoder @xcite .
we observe that the fer performance can be improved significantly by increasing the number of em iteration .
it is possible to achieve @xmath324 fer in expense of 4db of @xmath69 comparing to no phase noise scenario .
however , the error floor still occurs .
it is concluded that the number of the decoder iterations of the em - based algorithm , @xmath297 , should be tuned to a sufficiently large number to be able to track time varying phase noise .
in chapter [ chap : siso - model ] , phase noise tracking is performed by a hard decision directed ekf for the uncoded siso system .
numerical results shows that the ekf is able to track slowly time varying phase noise processes .
unsurprisingly , the ber degradation reaches large values when the phase noise innovation variance is high .
additionally , in chapter [ chap : siso - model ] , the problem of joint phase noise estimation and detection in a coded siso system with ldpc codes is discussed . the em - based algorithm proposed in @xcite is modified and analytically derived .
two estimators , a soft decision directed ks - mla and a soft decision directed eks proposed in @xcite are applied to carry out the maximization step of the em - based algorithm . in @xcite ,
the ks - mla is claimed to have superior performance than the eks .
however , numerical results in chapter [ chap : siso - model ] indicates that when phase noise over the frame reaches very large values , i.e. , in the case of large block length and/or phase noise innovation variance , the performance of the ks - mla degrades significantly and the eks performs better than the ks - mla .
the ks - mla removes the data dependency by multiplying the observed signal with the soft decision symbol . as a result
, the performance of the ks - mla degrades faster than the eks when the soft decisions are less reliable .
for instance , for the same phase noise innovation variance , the ekfs performs better than the ks - mla when the constellation density increases . a trick to increase the algorithm speed
is also discussed and shown to decrease the overall complexity required for convergence . in chapter [ chap : mimo - model ] , a low complexity hard decision directed ekf is derived and applied to an uncoded mimo system .
simulation results show that the ekf performs close to the synchronized system in the case of slowly time varying phase noise process .
the problem of joint estimation of the time varying phase noise and data detection for a los - mimo system using bit interleaved coded modulation and ldpc codes is also discussed . an iterative em - based receiver to perform code - aided synchronization is proposed .
a new low complexity soft - decision directed ekfs is derived and embedded into the em - based algorithm .
the performance of the system is investigated in terms of the ber and the fer .
estimation accuracy of the phase noise parameters is presented with average mse curves .
computational complexity of the system is also discussed .
computer simulations show that the number of bit errors does not decrease at each em algorithm iteration if the algorithm fails to converge .
instead , the fer performance is shown to decrease at each em iteration .
simulation results demonstrate that the proposed em - based algorithm estimates and compensates the time varying phase noise with a small degradation of the performance for a wide range of phase noise innovation variances
. however , an error floor occurs at high signal - to - noise ratio levels . to reduce the error floor and to track the phase noise process with large innovation variance decoding performance can be increased by setting the number of decoder iterations to larger values .
however , the achieved performance gain is not significant comparing to the introduced complexity . on the other hand ,
coding rate can be reduced further to protect data bits , yielding better soft decisions .
the fer performance can be significantly improved by utilizing low rate ldpc codes at the expense of a decrease in throughput . as a result
, the ekfs can be applied to a coded los - mimo system for the wide range of phase noise innovation variances .
the system parameters should be set to achieve the target performance . in chapter [ chap : mimo - model ] , the channel gains are assumed to be known at the receiver side .
the effects of the estimation errors of the channel gains on the performance of the em - based algorithm is not investigated .
the channel gains can be obtained by a conventional data - aided estimator .
additionally , the channel estimation of the block fading mimo systems can be embedded into the em - based algorithms and initial estimates can be improved at every em algorithm iteration .
therefore , this work can be extended to joint phase noise and channel estimation and data detection . in @xcite ,
a data - aided ls estimator a decision - directed wls estimator and a new decision - directed ekf is proposed .
these estimators can be used in the em - based algorithm to carry out the maximization step .
finally , statistics of the phase noise estimates can be used by taking into account in the decoding process .
a receiver employing factor graphs can be implemented and to track strong phase noise processes .
a. demir , a. mehrotra , and j. roychowdhury , `` phase noise in oscillators : a unifying theory and numerical methods for characterization , '' _ ieee trans .
circuits syst .
i , fundam .
theory appl .
_ , vol . 47 , no . 5 , pp .
655674 , may 2000 .
h. meyr , m. moeneclaey , and s. fechtel , _ digital communication receivers : synchronization , channel estimation , and signal processing_.1em plus 0.5em minus 0.4emnew york , ny , usa : john wiley & sons , inc . , 1997 .
s. godtmann , n. hadaschik , a. pollok , g. ascheid , and h. meyr , `` iterative code - aided phase noise synchronization based on the lmmse criterion , '' in _ ieee workshop on signal process .
advances in wireless commun .
_ , june 2007 , pp . 1 5 .
a. stefanov and t. duman , `` turbo - coded modulation for systems with transmit and receive antenna diversity over block fading channels : system model , decoding approaches , and practical considerations , '' _ ieee j. sel .
areas commun .
_ , vol . 19 , no . 5 , pp . 958968 , may 2001 .
j. boutros , f. boixadera , and c. lamy , `` bit - interleaved coded modulations for multiple - input multiple - output channels , '' in _ proc .
symposium on spread spectrum techniques and applications _ , vol . 1
2000 , pp .
123126 .
n. hadaschik , m. drpinghaus , a. senst , o. harmjanz , u. kufer , g. ascheid , and h. meyr , `` improving mimo phase noise estimation by exploiting spatial correlations , '' in _ proc .
conf . on acoustics , speech and signal process . _ , philadelphia , 2005 , pp .
833836 .
h. mehrpouyan , a. a. nasir , s. d. blostein , t. eriksson , g. karagiannidis , and t. svensson , `` joint estimation of channel and oscillator phase noise in mimo systems , '' _ ieee trans .
signal process .
_ , vol . 60 , no . 9 , pp
. 47904807 , sep .
g. golden , c. foschini , r. valenzuela , and p. wolniansky , `` detection algorithm and initial laboratory results using v - blast space - time communication architecture , '' _ electronics letters _ , vol .
35 , no . 1 ,
14 16 , jan . 1999 .
p. wolniansky , g. foschini , g. golden , and r. valenzuela , `` v - blast : an architecture for realizing very high data rates over the rich - scattering wireless channel , '' in _ ursi int .
symposium on signals , systems , and electronics _ , sep .
1998 , pp . 295300 .
v. tarokh , a. naguib , n. seshadri , and a. calderbank , `` space - time codes for high data rate wireless communication : performance criteria in the presence of channel estimation errors , mobility , and multiple paths , '' _ ieee trans .
_ , vol .
47 , no . 2 , pp . 199 207 , feb . 1999 . | new generation cellular networks have been forced to support high data rate communications .
the demand for high bandwidth data services has rapidly increased with the advent of bandwidth hungry applications . to fulfill the bandwidth requirement ,
high throughput backhaul links are required .
microwave radio links operating at high frequency bands are used to fully exploit the available spectrum .
generating high carrier frequency becomes problematic due to the hardware limitations .
non - ideal oscillators both at the transmitter and the receiver introduces time varying phase noise which interacts with the transmitted data in a non - linear fashion .
phase noise becomes a detrimental problem in digital communication systems and needs to be estimated and compensated . in this thesis receiver algorithms
are derived and evaluated to mitigate the effects of the phase noise in digital communication systems .
the thesis is organized as follows : in chapter [ chap : siso - model ] phase noise estimation in single - input single - output ( siso ) systems is investigated .
first , a hard decision directed extended kalman filter ( ekf ) is derived and applied to track time varying phase noise for an uncoded system .
next , the problem of phase noise estimation for coded siso system is investigated .
an iterative receiver algorithm performing code - aided turbo synchronization is derived using the expectation maximization ( em ) framework .
two soft - decision directed estimators in the literature based on kalman filtering , the kalman filter and smoother with maximum likelihood average ( ks - mla ) and the extended kalman filter and smoother ( eks ) , are evaluated .
low density parity check ( ldpc ) codes are proposed to calculate marginal a posteriori probabilities and to construct soft decision symbols .
error rate performance of both estimators , the ks - mla and the eks , are determined and compared through simulations .
simulations indicate that comparison on the performance of the existing estimators heavily depends on the system parameters such as block length and modulation order which are not taken into consideration in the literature . in chapter [ chap : mimo - model ]
the thesis focuses on phase noise estimation in multi - input multi - output ( mimo ) systems .
mimo technology is commonly used in microwave radio links to improve spectrum efficiency .
first , an uncoded mimo system is taken under consideration .
a low complexity hard decision directed ekf is derived and evaluated . a new mimo receiver algorithm that iterates between the estimator and the detector , based on the em framework for joint estimation and detection in coded mimo systems in the presence of time
varying phase noise is proposed .
a low complexity soft decision directed extended kalman filter and smoother ( ekfs ) that tracks the phase noise parameters over a frame is proposed in order to carry out the maximization step .
the proposed ekfs based approach is combined with an iterative detector that utilizes bit interleaved coded modulation and employs ldpc codes to calculate the marginal a posteriori probabilities of the transmitted symbols , i.e. , soft decisions .
numerical investigations show that for a wide range of phase noise variances the estimation accuracy of the proposed algorithm improves at every iteration .
finally , simulation results confirm that the error rate performance of the proposed em - based approach is close to the scenario of perfect knowledge of phase noise at low - to - medium signal - to - noise ratios . |
the observational evidence for an acceleration of the expansion of the universe is now overwhelming , although the precise cause of this phenomenon is still unknown ( see , e.g. , @xcite for recent reviews ) . in this concern , and besides the need for more accurate estimates of cosmological parameters , the current state of affairs also brings to light some other important aspects regarding the physics of the mechanism behind cosmic acceleration .
certainly , one of these aspects concerns the thermodynamical behavior of a dark energy - dominated universe , and questions such as `` what is the thermodynamic behavior of the dark energy in an expanding universe ? '' or , more precisely , `` what is its temperature evolution law ? '' must be answered in the context of this new conceptual set up .
another interesting aspect in this discussion is whether thermodynamics in the accelerating universe can place constraints on the time evolution of the dark energy and can also reveal some physical properties of this energy component .
the aim of this paper is twofold .
first , to derive physical constraints on the dark energy from the second law of thermodynamics and to deduce the temperature evolution law for a dark component with a general equation - of - state ( eos ) parameter @xmath6 ; second , to perform a joint statistical analysis involving current observational data together with the thermodynamic bounds on @xmath6 . to do that , we assume the following generalized formula for the time evolution of @xmath6 @xcite @xmath7 which recovers some well - known eos parameterizations in the following limits : @xmath8 where @xmath9 and @xmath10 stand for the dark energy pressure and energy density , respectively ( see also @xcite for other eos parameterizations ) .
the analyses are performed using one the most recent type ia supernovae ( sne ia ) observations , the nearby + sdss + essence + snls + hubble space telescope ( hst ) set of 288 sne ia discussed in ref .
@xcite ( which we refer to as sdss compilation ) .
we consider two sub - samples of this latter compilation that use salt2 @xcite and mlcs2k2 @xcite sn ia light - curve fitting method . along with the sne
ia data , and to help break the degeneracy between the dark energy parameters we also use the baryonic acoustic oscillation ( bao ) peak at @xmath11 @xcite and the current estimate of the cmb shift parameter @xmath12 @xcite .
we work in units where c = 1 . throughout this paper a subscript 0 stands for present - day quantities and a dot denotes time derivative .
let us first consider a homogeneous , isotropic , spatially flat cosmologies described by the friedmann - robertson - walker ( frw ) flat line element , @xmath13 , where @xmath14 is the cosmological scalar factor .
the matter content is assumed to be composed of baryons , cold dark matter and a dark energy component .
in such a background , the thermodynamic states of a relativistic fluid are characterized by an energy momentum tensor ( perfect - type fluid ) @xmath15 a particle current , @xmath16 and an entropy current , @xmath17 where @xmath18 is the usual projector onto the local rest space of @xmath19 and @xmath20 and @xmath21 are the particle number density and the specific entropy ( per particle ) , respectively @xcite .
the conservation laws for energy and particle number densities read @xmath22 @xmath23 where
semi - colons mean covariant derivative , @xmath24 is the scalar of expansion and the quantities @xmath9 , @xmath10 , @xmath20 and @xmath21 are related to the temperature @xmath25 trough the gibbs law : @xmath26 . from the energy conservation equation above , it follows that the energy density for a general @xmath6 component can be written as @xmath27\;.\ ] ] following standard lines ( see , e.g. , @xcite ) , it is possible to show that the temperature evolution law is given by @xmath28 where we have split the dark energy pressure as @xmath29 where @xmath30 . by combining the above equations , we also find that @xmath31\;,\ ] ] where we have used that @xmath32 , as given by the conservation of particle number density [ eq . ( [ nalpha ] ) ] . for parameterization p@xmath33 , shown in eq .
( [ pbeta ] ) , eq .
( [ temp ] ) can be rewritten as @xmath34\;,\ ] ] which reduces to the generalized stefan - boltzmann law for @xmath35 @xcite ( see also @xcite ) . from the above expressions , we confirm the results of ref .
@xcite ( for a constant eos parameter ) and find that dark energy becomes hotter in the course of the cosmological expansion since the its eos parameter must be a negative quantity .
a possible physical explanation for this behavior is that thermodynamic work is being done on the system ( see , e.g. , fig . 1 of ref .
in particular , for the vacuum state ( @xmath36 ) we obtain @xmath37 . to illustrate this behavior , fig .
1 shows the dark energy temperature as a function of the scale factor for @xmath38 ( p2 ) and @xmath39 ( p3 ) , with the dark energy density @xmath10 blowing up as @xmath40 at high-@xmath41 .
we , therefore , do not consider this parameterization in our analyses .
] by assuming arbitrary values of @xmath4 , @xmath5 and @xmath42 , where @xmath43 k. from this analysis , it is clear that an important point for the thermodynamic fate of the universe is to know how long the dark energy temperature will take to become the dominant temperature of the universe .
a basic difficulty in estimating such a time interval , however , is that the present - day dark energy temperature has not been measured , being completely unknown . 0.1 in by considering that the chemical potential for this @xmath6-fluid is null ( as occurs for @xmath44 )
, the euler s relation defines its specific entropy , i.e. , @xmath45 now , by combining the above equations , it is straightforward to show that the product @xmath46 , so that @xmath47 for a constant eos parameter , the above expression recovers some of the results of ref .
note also that the vacuum entropy is zero ( @xmath36 ) whereas for phantom dark energy ( @xmath48 ) , which violates all the energy conditions @xcite , the entropy assumes negative values being , therefore , meaningless ( for a discussion on the behavior of a phanton fluid with nonzero chemical potential , see @xcite .
see also @xcite for an alternative explanation in which the temperature of the phantom component takes negative values and @xcite for other thermodynamic analyses of dark energy ) .
two cases of interest arise directly from eq .
( [ entropy ] ) .
the case in which @xmath49 implies necessarily that @xmath50 for all the above parameterizations are imcompatible with the case @xmath51 . ] .
the second case is more interesting , with the @xmath6-fluid mimicking a fluid with bulk viscosity where the viscosity term is identified with the varying part of the dark energy pressure @xmath52 . for this latter case
, we note that the positiveness of @xmath53 implies that @xmath54 which clearly is not defined at @xmath55 , where @xmath56 . finally , by combining eqs .
( [ euler ] ) and ( [ entropy ] ) with the conservation of the particle number density shown earlier , we obtain from the second law of thermodynamics that @xmath57 or , equivalently , @xmath58 .
in this section , we combine the above physical constraints ( [ c1 ] ) and ( [ c2 ] ) with current observational data in order to impose bounds on the dark energy parameters .
we use one of the most recent sne ia data sets available , namely , the sdss compilation discussed in ref . @xcite .
this compilation comprises 288 sne ia and uses both salt2
@xcite and mlcs2k2 @xcite light - curve fitters ( see also @xcite for a discussion on these light - curve fitters ) and is distributed in redshift interval @xmath60 .
along with the sne ia data , and to help break the degeneracy between the dark energy parameters @xmath4 and @xmath61 we use the bao @xcite and shift parameters @xcite @xmath62 @xmath63 where @xmath64^{1/3}$ ] is the so - called dilation scale , defined in terms of the dimensionless comoving distance @xmath65 , @xmath11 and @xmath66 . in our analyses , we minimize the function @xmath67 , which takes into account all the data sets mentioned above and marginalize over the present values of the matter density @xmath68 and hubble parameters @xmath69 . figures 2 and 3 show the main results of our joint analyses .
we plot contours of @xmath70 in the parametric space @xmath4 - @xmath5 for p2 and p3 , respectively .
the light gray region displayed in the plots stands for the physical constraint ( [ c1 ] ) . since this inequality is a function of time , the region is plotted by assuring its validity from @xmath71 up to today at @xmath55 .
the resulting parametric space , when all the observational data discussed above are combined with the constraints ( [ c1 ] ) and ( [ c2 ] ) , corresponds to the small hachured area right below the @xmath35 line .
these results clearly illustrate the effect that the thermodynamic bounds discussed in the previous section may have on the determination of the dark energy eos parameters .
in particular , we note that the resulting allowed regions are even tighter for the logarithmic parameterization p2 than for p3 ( cpl ) . since the salt2 compilation allows for more negative values of @xmath4 , the joint constraints involving this sne ia sub - sample are also more restrictive ( figs .
2b and 3b ) . for completeness
, we display in table i the changes in the 2@xmath21 estimates of @xmath4 and @xmath5 due to the thermodynamic bounds ( [ c1 ] ) and ( [ c2 ] ) . [ n - table ]
lccc + test & & @xmath4 & @xmath5 + + sne ia ( mlcs2k2 ) .................... & p2 & @xmath72 & @xmath73 + sne ia ( mlcs2k2)@xmath74 + t ] .......... & p2 & @xmath75 & @xmath76 + sne ia ( salt2)@xmath74 ........................ & p2 & @xmath77 & @xmath78 + sne ia ( salt2)@xmath74 + t@xmath79 ............... & p2 & @xmath80 & @xmath81 + sne ia ( mlcs2k2)@xmath74 .................... & p3 & @xmath82 & @xmath83 + sne ia ( mlcs2k2)@xmath74 + t@xmath79 .......... & p3 & @xmath84 & @xmath85 + sne ia ( salt2)@xmath74 ........................ & p3 & @xmath86 & @xmath87 + sne ia ( salt2)@xmath74 + t@xmath79 ............... & p3 & @xmath88 & @xmath89 +
in spite of its fundamental importance for an actual understanding of the evolution of the universe , the relevant physical properties of the dominant dark energy component remain completely unknown . in this paper
we have investigated some thermodynamic aspects of this energy component assuming that its constituents are massless quanta with a general time - dependent eos parameter @xmath6 .
we have discussed its temperature evolution law and derived constraints from the second law of thermodynamics on the values of @xmath4 and @xmath5 for a family of @xmath6 parameterizations given by eq .
( [ pb ] ) .
when combined with current data from sne ia , bao and cmb observations , we have shown that such constraints provide very restrictive limits on the parametric space @xmath4 - @xmath5 ( see figs . 2 and 3 ) . finally , it is also worth mentioning that in the present analysis we have assumed that the chemical potential @xmath90 for the @xmath6-fluid representing the dark energy is null . a more general analysis relaxing this condition ( @xmath91 ) is currently under preparation and will appear in a forthcoming communication .
b. ratra & m. s. vogeley , pasp * 120 * , 235 ( 2008 ) ; r. r. caldwell & m. kamionkowski , ann .
nucl . part .
sci . * 59 * , 397 ( 2009 ) ; a. silvestri & m. trodden , rept .
. phys . *
72 * , 09690 ( 2009 ) ; m. sami , curr .
sci . * 97 * , 887 ( 2009 ) .
y. wang and p. m. garnavich , astrophys .
j. * 552 * , 445 ( 2001 ) ; c. r. watson and r. j. scherrer , phys .
d * 68 * , 123524 ( 2003 ) ; p.s .
corasaniti et al . , phys .
d * 70 * , 083006 ( 2004 ) ; v. b. johri , astro - ph/0409161 ; y. wang and m. tegmark , phys .
* 92 * , 241302 ( 2004 ) ; h. k. jassal , j. s. bagla , and t. padmanabhan , mon . not .
. soc . * 356 * , l11 ( 2005 ) ; e. m. barboza jr . and j. s. alcaniz , phys .
b * 666 * , 415 ( 2008 ) .
m. m. phillips , astrophys . j. * 413 * , l105 ( 1993 ) ; a. g. riess , w. h. press , and r. p. kirshner , astrophys .
j. * 438 * , l17 ( 1995 ) ; s. jha , a. g. riess , and r. p. kirshner , astrophys . j. * 659 * , 122 ( 2007 ) .
g. izquierdo and d. pavon , phys .
b * 633 * , 420 ( 2006 ) ; n. bilic , fortsch .
phys . * 56 * , 363 ( 2008 ) ; y. s. myung , phys .
b * 671 * , 216 ( 2009 ) ; e. n. saridakis , p. f. gonzalez - diaz and c. l. siguenza , class .
. grav .
* 26 * , 165003 ( 2009 ) . | a significant observational effort has been directed to unveil the nature of the so - called dark energy .
however , given the large number of theoretical possibilities , it is possible that such a task can not be performed on the basis only of the observational data . in this article
we discuss some thermodynamic properties of this energy component assuming a general time - dependent equation - of - state parameter @xmath0 , where @xmath1 and @xmath2 are constants and @xmath3 may assume different forms .
we show that very restrictive bounds can be placed on the @xmath4 - @xmath5 space when current observational data are combined with the thermodynamic constraints derived . |
diamond cavity optomechanical devices are an attractive platform for controlling interactions between light , vibrations , and electrons that underly future hybrid quantum technologies @xcite .
their potential arises from diamond s exceptional mechanical and optical properties @xcite combined with its ability to host color centers such as the nitrogen - vacancy ( nv ) whose electron spins are excellent qubits that can be manipulated by local mechanical strain fields @xcite .
recently , piezoelectric actuation of bulk @xcite and nanomechanical @xcite diamond resonators has been used to demonstrate phononic spin control .
cavity optomechanics @xcite harnesses optical forces in place of piezoelectric actuation , allowing coherent phonon state manipulation @xcite of ghz frequency mechanical resonators with quantum limited sensitivity @xcite .
these phonons can be made resonant with nv center electron spin transitions that are central to proposals for spin
spin entanglement @xcite , spin phonon state transfer @xcite , spin mediated mechanical normal mode cooling @xcite , and photon - phonon - spin coupling @xcite .
additionally , the relatively small thermal occupancy and mechanical dissipation of ghz diamond devices , combined with diamond s ability to support strong optical fields due its large electronic bandgap , make them an ideal system for the cavity optomechanical backaction cooling and study of mechanical resonators in their quantum ground state @xcite .
development of cavity optomechanical devices from single - crystal diamond has been limited due to challenges associated with fabricating mechanically isolated structures from bulk diamond chips . while initial development of diamond optomechanical devices used nanocrystalline diamond @xcite , single crystal diamond material promises lower mechanical dissipation @xcite and the ability to host highly coherent nv centers @xcite . here
we report demonstration of a single - crystal diamond cavity optomechanical system for the first time .
this system is based on a microdisk device geometry that has been used in a wide range of cavity optomechanics experiments implemented in more conventional semiconductor and dielectric materials @xcite .
microdisks are desirable owing to their simple geometry , strong optomechanical coupling between high frequency mechanical resonances and low loss optical modes , and intrinsic ability to simultaneously support optical modes over the entire transparency window of the device material @xcite .
the microdisk system studied here , an example of which is shown in fig .
[ fig : sem](a ) , supports optical modes at visible and telecommunication wavelengths ( @xmath7 thz ) that interact via radiation pressure with ghz frequency mesoscopic mechanical resonances of the structure .
we find that these resonances have a record combination of high @xmath8 and low mechanical dissipation ( @xmath9 ) compared to other mechanical resonators operating in ambient temperature and pressure , and that their @xmath10 hz product is sufficiently high to satisfy the minimum criteria for single phonon coherent behaviour @xcite .
the microdisk optical modes have low dissipation ( @xmath11 ) , and owing to the negligible nonlinear absorption in diamond at telecom optical frequencies , they can support intracavity photon number @xmath12 without degrading @xmath13 . in combination , this allows realization of optomechanical cooperativity , @xmath14 , large enough ( @xmath15 ) for coherent photon - phonon coupling @xcite , where @xmath16 is the single photon optomechanical coupling rate of the device and describes the expected shift in cavity optical frequency due to the mechanical zero point motion of the microdisk .
these devices operate on the border of the sideband resolved regime ( @xmath17 ) , enabling radiation pressure backaction excitation of mechanical self - oscillations with @xmath18 amplitude .
the accompanying stress fields are strong enough to drive diamond color center spin transitions with a single phonon - spin coupling rate that is predicted to exceed that of previously studied mhz frequency nanomechanical structures @xcite , despite having orders of magnitude higher @xmath8 and smaller phonon amplitude , owing to the localized nature of the microdisk mechanical resonances .
in addition , the ability of the microdisks to support optical modes at visible wavelengths is compatible with resonant coupling to nv center optical transitions @xcite , as well as operation in fluid environments of interest for sensing applications @xcite .
there has been significant recent progress in fabrication of mechanically isolated single - crystal diamond devices , including demonstrations of suspended high-@xmath19 nanomechanical resonators @xcite and high-@xmath13 micro- and nanocavities @xcite .
these structures have been created using diamond membrane thinning @xcite , plasma angled - etching @xcite , and plasma undercutting @xcite fabrication techniques , with the latter two approaches allowing patterning of devices from bulk diamond chips .
here we use plasma undercutting to fabricate single crystal diamond cavity optomechanical devices @xcite .
these devices were fabricated from an optical grade , chemical vapor deposition ( cvd ) grown @xmath20-oriented scd substrate supplied by element six .
the polished substrates were first cleaned in boiling piranha , and coated with @xmath21 400 nm of plasma - enhanced chemical vapor deposition ( pecvd ) si@xmath22n@xmath23 as a hard mask . to avoid charging effects during electron beam lithography ( ebl ) , @xmath21 5 nm of ti was deposited on the si@xmath22n@xmath23 layer , before coating the sample with the zep 520a ebl resist .
the developed pattern was transferred to the hard mask via inductively coupled reactive ion etching ( icprie ) with c@xmath23f@xmath24/sf@xmath25 chemistry .
the remaining ebl resist was removed with a 6 minute deep uv exposure ( @xmath26 at 254 nm ) followed by a 2 minute soak in remover pg , while the remaining ti was removed by the subsequent etch steps .
the anisotropic icprie diamond etch was performed using o@xmath27 , followed by deposition of @xmath21 250 nm of conformal pecvd si@xmath22n@xmath23 as a sidewall protection layer .
the bottom of the etch windows were then cleared of si@xmath22n@xmath23 using a short icprie c@xmath23f@xmath24/sf@xmath25 etch .
this was followed by a zero rf power o@xmath27 rie diamond undercut etch to partially release the devices .
lastly , the si@xmath22n@xmath23 layer was removed using a wet - etch in 49@xmath28 hf , and the devices were cleaned again in boiling piranha .
the devices studied here have diameters of @xmath29 to @xmath30 and average thickness @xmath31 . as evident from the image in fig .
[ fig : sem](a ) , devices are fabricated with a process optimized to minimize the waist of the pedestal supporting the microdisk , reducing it to @xmath32 , where the waist is defined as the smallest point of the pedestal .
the microdisk thickness , which will be reduced in future work to enhance confinement , is determined by the interplay between the inward and upward etch rates of the quasi - isotropic undercut , together the initial anisotropic etch depth .
the undercut time was chosen to optimize the pedestal waists of the @xmath33 diameter disks studied here . a longer
undercut would allow the study of larger diameter microdisks ( 6 @xmath34 to 8 @xmath34 ) present on the chip , which would in turn possess a smaller thickness than the structures studied here .
the devices were characterized by monitoring the transmission of a dimpled optical fiber taper @xcite evanescently coupled to the microdisk and input with light from tunable diode lasers ( new focus velocity ) with wavelengths near 1530 nm or 637 nm .
for the 1530 nm measurements , the output of the fiber taper was monitored by both low- and high - bandwidth photoreceivers ( newport 1621 and 1554-b , respectively ) , and a calibrated optical power meter ( newport 2936-r ) .
figure [ fig : sem](b ) shows typical @xmath35 when the fiber taper is evanescently coupled to devices with diameters of @xmath29 and @xmath36 and the wavelength @xmath37 of the 1530 nm tunable laser is scanned across microdisk modes at @xmath38 . here
@xmath39 is the average transmission measured by the low - bandwidth photodetector over a timescale long compared to @xmath40 .
these measurements reveal resonant coupling to modes with loaded @xmath41 ( intrinsic @xmath42 ) , and a degree of doublet structure that depends on the internal backscattering of a given device .
maximizing @xmath13 , and thereby minimizing @xmath43 , is important for achieving the aforementioned regime allowing coherent photon
phonon coupling ( @xmath44 ) .
the ability of these devices to support modes over a wide wavelength range is demonstrated in fig.[fig : sem](c ) , where the 637 nm tunable laser was used to probe a mode with high-@xmath45 .
this is promising for applications involving nv center optical transitions in this wavelength range @xcite .
these devices have a predicted radiation loss limited @xmath46 at both 1550 nm and 637 nm wavelengths ; @xmath43 is currently limited by surface roughness and linear absorption . in previous work
the microdisk pedestal size was observed to limit @xmath13 for insufficient relative undercut . in the devices studied here scattering due to the non
cylindrical pedestal shape is predicted to dominate the contribution to @xmath43 from the pedestal @xcite .
the lower @xmath13 observed at 637 nm can be attributed in part to sub optimal fiber taper positioning @xcite and an increased sensitivity to surface scattering @xcite . to probe optomechanical coupling within the microdisks , time ( @xmath47 ) dependent transmission fluctuations @xmath48 of @xmath49 light
were monitored using a real time spectrum analyzer ( tektronix rsa5106a ) .
excitations of the microdisk mechanical resonances modulate @xmath38 , resulting in a dispersive optomechanical transduction of mechanical motion to an optical signal @xmath50 that can then be observed in the measured electronic power spectrum @xmath51 . here
@xmath52 is the average power transmitted to the photoreceiver .
figure [ fig : sem](d ) shows typical spectra when @xmath37 is tuned near the point of maximum transduction of the modes in fig .
[ fig : sem](b ) .
resonances near @xmath53 are observed , corresponding to optomechanical transduction of the thermomechanical motion of the fundamental radial breathing mode ( rbm ) of the microdisks .
the predicted displacement of the rbm calculated using finite element simulations ( comsol ) is shown in the inset to fig .
[ fig : sem](a ) .
the simulated @xmath1 of the rbm for varying microdisk diameter was found to be within 10% of observed values .
these measurements were conducted at low input power @xmath54w to avoid optomechanical backaction effects discussed below . here an erbium doped fiber amplifier ( edfa : pritel lnhpfa-30 ) was used on the output side of the fiber taper to boost the optical signal prior to photodetection to a level just below the detector saturation power ( @xmath55 mw ) . the microdisk pedestal can significantly affect the rbm properties , and minimizing its waist size is important in order to maximize @xmath19 and reach @xmath44 . in previously studied diamond microdisks with @xmath56 m pedestal waists @xcite , transduction of mechanical modes was not observed . as shown in fig .
[ fig : sem](a ) , the devices used here for cavity optomechanics have significantly smaller waists , e.g. , the @xmath29 diameter microdisks have waist @xmath32 . figure [ fig : sem](d ) shows that when microdisk diameter , and as a result , pedestal waist are increased to @xmath36 and @xmath57 , respectively , @xmath19 is found to decrease from @xmath58 to @xmath59 .
mechanical resonances are not observed in devices with pedestal diameter @xmath60 nm .
this indicates that @xmath19 for these devices is limited by clamping loss @xcite .
the hourglass shape of the pedestals obtained for the @xmath61 optical grade diamond samples used here limits the minimum size of the pedestal where it connects to the microdisk , and may result in increased dissipation .
given the crystal plane selective nature of the diamond undercut @xcite , fabricating microdisks from samples with alternate crystal orientation such as @xmath62 may alleviate this limitation .
additionally , operation in vacuum , where viscous air damping can be avoided @xcite , and at low temperature @xcite would allow a decrease in dissipation , as the total @xmath19 is given by @xmath63 , where the @xmath64 represent the quality factor due to each damping mechanism . despite present limitations , the demonstrated devices have @xmath65 hz , that is larger than all previously studied cavity optomechanical systems operating in ambient conditions @xcite .
a comparison of some of the highest @xmath66 products for optomechanical systems observed in ambient , cryogenic , and low pressure environments is shown in supplement 1 , section 3 .
this figure of merit is critical for cavity optomechanical mass spectroscopy @xcite and low phase noise oscillators @xcite . within the context of quantum optomechanics
, this product satisfies a key minimum requirement for room temperature studies of single phonon coherence by over an order of magnitude : @xmath67 , where @xmath68 is the room temperature phonon population of the rbm , ensuring that thermal decoherence is slow compared to a mechanical oscillation @xcite . by satisfying this condition , cooling to the quantum ground state from room temperature
should also be possible @xcite . to investigate the response of the cavity optomechanical transduction , @xmath69 was monitored while @xmath37 was scanned across @xmath38 .
figure [ fig : optomechanics](a ) shows the resulting measurement of @xmath70 for the microdisk in fig .
[ fig : sem](a ) , clearly illustrating that optomechanical transduction is only observable when @xmath37 is tuned in the vicinity of @xmath71 . in this measurement
the edfa was connected to the input side of the fiber taper , resulting in maximum @xmath72 and @xmath73 mw , where @xmath74 is the optical power dropped into the microdisk mode . from the thermo
optic coefficient of diamond we estimate that the shift of 400 pm from the cold cavity @xmath38 , as seen in figure [ fig : optomechanics](a ) , corresponds to a change in device temperature @xmath75 k. comsol simulations that take into account the reduced thermal conductivity coefficient of the @xmath76 nm diameter pedestal compared to bulk @xcite confirm that an absorbed power of @xmath77 of @xmath74 reproduces the temperature shift observed here ( see supplement 1 , section 1 ) .
although this temperature increase is not desirable for quantum optomechanics applications , it is considerably smaller than the expected increase for a similar si device . assuming a similar absorption rate and identical device geometry , a silicon device with a silicon or silicon dioxide pedestal would result in @xmath78 k or 450 k , respectively , where a modified thermal conductivity also applies for the silicon in the pedestal @xcite .
it is expected that the rate of linear absorption observed here , which corresponds to @xmath79 , can be reduced through improvements to processing , as diamond devices with @xmath80 have been reported elsewhere @xcite .
the influence of cavity optomechanical backaction @xcite on the dynamics of the mechanical resonator is analyzed in figs .
[ fig : optomechanics](b , c ) .
changes @xmath81 and @xmath82 to @xmath83 and @xmath84 , respectively , were measured as a function of source cavity detuning @xmath85 .
their strong dependence on @xmath86 clearly indicates that the mechanical dynamics are affected by the intracavity field .
we were prevented from measuring significant @xmath81 and @xmath82 for red - detuned wavelengths ( @xmath87 ) due to the thermal bistability present in our system for large @xmath74 , as shown in fig .
[ fig : optomechanics](a ) . as such
, this study concentrated on blue - detuned wavelengths ( @xmath88 ) , however implementation of cavity stabilization techniques @xcite may allow this limitation to be overcome in the future , enabling more effective investigations of cavity sideband cooling @xcite , and optomechanically induced transparency @xcite .
determining @xmath86 for each data point in this analysis required accounting for the dependence of @xmath89 on @xmath90 due to the thermo - optic effect . for a given operating @xmath91 ,
@xmath89 was predicted from @xmath90 ( see supplement 1 , section 1 ) , where @xmath92 was determined from @xmath93 , @xmath94 , @xmath95 and measurements of loss through the fiber taper and other elements of the apparatus .
to quantitatively investigate the role of radiation pressure on the mechanical resonance dynamics , the observed @xmath96 was fit to the expected cavity optomechanical damping rate @xcite , with the single photon optomechanical coupling rate @xmath97 as the only free parameter .
using this method , the fit shown in fig . [ fig : optomechanics](b ) was obtained for @xmath98 , with an associated 95% confidence interval of @xmath99 khz .
errors bars for each @xmath100 data point in fig .
[ fig : optomechanics](b ) represent the 95% confidence interval of fits used to extract @xmath83 from @xmath69 . the large uncertainty as well as the discrepancy between the measured and predicted values when @xmath101 or @xmath102 are due to the low signal to noise of @xmath69 in these regions .
this low signal to noise of @xmath69 also prohibited measuring the optomechanical response for @xmath103 .
the fit value for @xmath97 has good agreement with @xmath97 predicted from comsol calculations that include both moving boundary ( mb ) and photoelastic ( pe ) contributions @xcite .
the predicted @xmath97 is dependent on the spatial overlap of the optical field and mechanical displacement profile , which varies for each optical mode .
we find that the fit value of @xmath97 most closely agrees with the predicted coupling rate to the second order radial tm like mode , with @xmath104 and @xmath105 . in comparison ,
@xmath106 ( @xmath107 ) and @xmath108 to @xmath109 ( @xmath110 to @xmath111 ) for the fundamental tm ( te ) mode of the microdisk .
this is consistent with measurements of the mode polarization in the fiber taper that indicated that the microdisk mode studied here is tm polarized .
note that the stated uncertainty in the predicted @xmath112 is due to variations in reported pe coefficients of single - crystal diamond @xcite .
figure [ fig : optomechanics](c ) shows a similar analysis of @xmath113 , indicating that @xmath8 is softened by over @xmath114 by the intracavity field .
this shift is due to both optomechanical dynamical backaction and static thermal effects . for the operating regime and devices used here
, dynamic thermal effects are expected to be below 5% of the optomechanical radiation pressure dynamical backaction effects , and can be neglected @xcite . however , static thermal effects are significant .
heating of the microdisk for large @xmath74 results in both thermal expansion and a change in young s modulus , resulting in a shift to @xmath8 @xcite .
this effect is linear in @xmath74 , assuming that @xmath13 is independent of power , i.e. , nonlinear absorption is small . to compare the measured @xmath113 with theory , we used a model that includes radiation pressure induced optomechanical dynamic backaction @xcite and a static heating term linearly proportional to @xmath74 : @xmath115 the resulting fit of to the measured @xmath116 is shown in fig .
[ fig : optomechanics](c ) to have close agreement .
notably , this model reproduces the kink in @xmath113 where the amplitude of the optomechanical contribution reaches a maximum and changes sign .
this fit was obtained with @xmath97 fixed to the value extracted from the analysis of @xmath100 in fig .
[ fig : optomechanics](b ) , and with @xmath117 as a fitting parameter . at higher power the microdisk optomechanical dynamics can be dramatically modified .
figure [ fig : self - oscillation ] shows the behavior of the microdisk rbm when the input power to the fiber taper is increased sufficiently for @xmath74 to reach 13 mw .
this elevated power level corresponds to an intracavity photon number @xmath118 , and an optomechanical cooperativity @xmath119 for the device shown in fig .
[ fig : sem](a ) .
this @xmath120 exceeds all previously reported values for devices operating in ambient conditions @xcite .
for such large @xmath74 , if the input @xmath37 is appropriately blue detuned from @xmath38 , it is possible for @xmath121 , resulting in self - oscillation of the microdisk rbm .
operation in this regime results in large dynamical strain within the microdisk , offering a potential path for achieving for large nv spin - phonon coupling .
we predict the strain achievable in our devices from measurements as follows .
the microdisk mechanical response in the transition from thermal motion to self - oscillation is shown in fig .
[ fig : self - oscillation](a ) , which displays @xmath122 for varying @xmath74 , with @xmath37 tuned to the value @xmath123 where @xmath124 is maximum . as @xmath74
is increased the mechanical resonance is observed to narrow and increase in amplitude , suggestive of the onset of self - oscillations , also referred to as phonon lasing .
this is more clearly illustrated in fig .
[ fig : self - oscillation](b ) , which shows the normalized spectrum @xmath125 for varying @xmath74 . here
@xmath125 has been obtained by normalizing @xmath122 with the power dependent transduction gain , such that the area under @xmath125 represents the mechanical energy of the rbm , i.e. @xmath126 is constant with respect to @xmath94 in absence of optomechanical backaction ( see supplement 1 , section 2 ) . at low power , this mechanical energy is dominantly from the thermal bath , and is predicted from the equipartition theorem to correspond to oscillation amplitude @xmath127 .
figures [ fig : self - oscillation](c ) and ( d ) show that for large @xmath74 a maximum @xmath128 is reached , likely limited by nonlinearity of the material .
the corresponding predicted stress maximum associated with the self - oscillations , also shown in figs .
[ fig : self - oscillation](c ) and ( d ) , is @xmath129 .
stress values were determined from @xmath130 and finite element simulations of the rbm displacement field shown in fig .
[ fig : sem](a ) , and the maximum is predicted to occur at the center the microdisk top surface , as shown by the plots in figs .
[ fig : self - oscillation](e , f ) .
the corresponding maximum strain is @xmath131 .
the self - oscillation threshold behaviour can be quantitatively analyzed by extracting the mechanical displacement amplitude @xmath130 as a function of @xmath74 from @xmath126 . this is shown figs .
[ fig : self - oscillation](c , d ) for two similarly sized microdisks . in each case , a clear threshold is observed . since these microdisks have different @xmath19 and @xmath13 ( see fig .
[ fig : self - oscillation ] caption ) , their threshold power @xmath132 differ . for devices close to the sideband resolved regime , the optimal detuning for self - oscillation to occur is @xmath133 , and @xmath132 is given by @xcite @xmath134\ ] ] where @xmath135 is the optomechanical coupling coefficient , and @xmath136 and @xmath137 are the intrinsic and fiber taper loaded optical decay rates , respectively . here
@xmath138 is the mechanical zero point motion amplitude . for the devices studied here , @xmath139 , as calculated from the rbm effective mass @xmath140 , predicted by the finite element simulated displacement field shown in fig .
[ fig : sem](a ) @xcite .
the observed @xmath141 and @xmath142 , for the devices in figs .
[ fig : self - oscillation](c ) and [ fig : self - oscillation](d ) respectively , are above the predicted values of @xmath143 and @xmath144 obtained from assuming @xmath97 is given by the fits in fig .
[ fig : optomechanics ] . this disagreement could be related to uncertainty in @xmath43 given that @xmath132 scales to the fourth power of this quantity , interplay between doublets that is ignored by , and uncertainty in @xmath86 inferred from the cavity response in the presence of thermo - optic dispersion .
the potential of these devices for hybrid spin - optomechanics applications can be measured by the predicted strain coupling rate @xmath145 between a single phonon of the microdisk rbm and a single diamond nv center electron spin .
the maximum zero point motion strain of the rbm is @xmath146 , and we estimate @xmath147 for a negatively charged nv@xmath148 center electron spin optimally located @xmath149 below the top surface of the device in fig .
[ fig : self - oscillation](c ) , exceeding the highest rate demonstrated to date @xcite . here
@xmath150 ghz is the strain susceptibility of the ground state spin @xcite . when the rbm is self - oscillating as in fig .
[ fig : self - oscillation](c ) , the predicted coupling rate is @xmath151 @xcite .
this is comparable to coupling rates achieved in piezoelectric actuated nanomechanical @xcite and bulk devices @xcite .
the longest room - temperature ground state spin decoherence time ( @xmath152 ) observed to date in isotopically engineered single - crystal diamond is 1.8 ms @xcite , while typical @xmath152 values in nanostructures are on the order of 100 @xmath56s @xcite .
additionally , dynamical decoupling schemes can be utilized to extend this time , as @xmath153 ms has been observed at low temperature @xcite .
spin control should be possible provided @xmath154 , which is the case for these devices .
the ghz frequency of the rbm enables low room temperature phonon occupation , relevant for cooling to the quantum ground state from room temperature @xcite .
this also enables access to larger energy spin transitions @xcite than possible using previously demonstrated diamond nanomechanical resonators
. this may be particularly important for future studies of phonon coupling to the nv@xmath148 center excited state manifold , which could achieve single phonon coupling rates close to a mhz due to the @xmath155 times larger strain susceptibility of the excited states @xcite .
this is promising for implementing fully quantum photon - phonon - spin interfaces , and for proposals of spin - mediated cooling of nanomechanical resonators @xcite . in the samples under study , we expect to find nvs optimally coupled to the rbm since the nitrogen concentration for this diamond sample ( @xmath21 ppm , corresponding to a number density of @xmath156 ) results in high - concentration nv ensembles .
however , future studies with higher purity samples may require nv implantation to optimally locate nvs @xmath21 50 nm below the device surface . additionally , due to minimal coupling of fluorescence from an nv centre located at the centre of the disk to the optical modes , free space collection would most likely be required . however , use of higher order radial breathing modes could allow for greater spatial overlap of the strain and electromagnetic field maxima @xcite , allowing for more efficient fiber based excitation and collection .
future improvements of @xmath13 to values above @xmath157 and approaching @xmath158 should be possible @xcite , enabling ultralow self - oscillation threshold @xcite and operation deep in the sideband resolved regime required for optomechanical ground state cooling @xcite .
operating in vacuum , at low temperature , and using devices fabricated from high - purity electronic grade diamond may allow further increases in @xmath19 @xcite , boosting the achievable photon
phonon cooperativity @xmath120 and @xmath66 product .
similarly , reducing the microdisk diameter may increase @xmath120 through enhanced @xmath97 , while also increasing @xmath8 .
simulations predict that diameters close to @xmath159 are possible before radiation loss limits @xmath160 ; such devices would have @xmath161 khz and @xmath162 .
finally , using electronic grade diamond material and investigating processing techniques to reduce surface state absorption may decrease optical absorption and allow larger @xmath90 before device heating becomes significant .
in conclusion , we have shown that cavity optomechanical devices can be realized from single - crystal diamond , with record high ambient condition optomecanical cooperativity , @xmath163 , and @xmath66 product of @xmath164 .
these devices are a promising testbed for ambient condition coherent optomechanics experiments , e.g. ground state cooling @xcite , optomechanically induced transparency @xcite and phonon mediated wavelength conversion @xcite , as well as studies in quantum information science @xcite , and hybrid quantum systems involving light , phonons , and diamond nv center spins @xcite .
we have also shown that the microdisks demonstrated here support high-@xmath13 optical modes at wavelengths near resonant with the @xmath165 optical transition of nv centers , further enhancing their potential for photon - phonon - spin coupling experiments .
we note that , in parallel to this work , burek et al . have demonstrated cavity optomechanics in single
crystal diamond optomechanical crystals fabricated by a faraday cage angled - etching technique @xcite .
national research council canada ( nrc ) ; natural sciences and engineering research council of canada ( nserc ) ; national institute for nanotechnology ( nint ) , canada foundation for innovation ( cfi ) ; alberta innovates technology futures ( aitf ) .
the authors would like to thank a. c. hryciw for his assistance with this project .
here we outline the process for extracting the power dependent detuning , @xmath86 .
this process follows carmon et al .
@xcite , beginning with the expression for the shifted cavity resonance wavelength as a function of temperature , in thermal equilibrium @xmath166\,,\\ & = \lambda_\text{o}\left[1+a \delta \text{t}\right]\label{eq : temp_shift}.\end{aligned}\ ] ] this expression is obtained by considering thermal expansion of the cavity , determined by the thermal expansion coefficient @xmath167 , and the thermo - optic effect , which shifts the refractive index @xmath168 with temperature @xmath169 . here
@xmath170 and @xmath171 are geometric factors accounting for the optical mode overlap with the changing @xmath168 and volume , respectively .
lumped constant @xmath172 describes the net thermo - optic dispersion of the cavity mode . using the room temperature single crystal diamond values of @xmath173 and @xmath174
we can estimate the change in temperature of the cavity as @xmath175\cdot \frac{1}{a}.\ ] ] the shift of @xmath176 pm , as seen in fig .
2(a ) of the main text , corresponds to a change in device temperature @xmath75 k. in this device the diamond forming the @xmath76 nm diameter pedestal has a significantly smaller thermal conductivity than that of bulk diamond ( @xmath177 ) , reaching values @xmath178 for nanowires @xmath179 nm in diameter @xcite . in order to confirm that the cavity temperature shift predicted by was reasonable for our system we performed finite element comsol simulations to estimate @xmath180 , including the modified thermal conductivity for the pedestal , as shown in fig.[fig : delta_t ] for varying pedestal widths .
fig.[fig : delta_t ] indicates that for a pedestal width of @xmath181 nm , and corresponding diamond thermal conductivity of @xmath182 a shift of 50 k is expected when @xmath183w , where @xmath184 is the total power absorbed by the cavity .
this corresponds to an optical absorption rate , @xmath185 312 mhz , which is @xmath186 of the total cavity decay rate , @xmath187 .
a linear relationship between @xmath180 and @xmath184 is observed for the pedestal thicknesses studied here . for varying pedestal widths .
here the total heat flow to the device is given by @xmath188 , where @xmath189 is the volume defined by the outer edge of the microdisk , with @xmath190 .
each line represents a linear line of best fit to @xmath180 as a function of @xmath184 .
[ fig : delta_t ] to convert ( [ eq : temp_shift ] ) to a form that depends on the experimentally measured , normalized cavity transmission @xmath39 , we treat the microdisk as being in thermal equilibrium with its environment such that @xmath191 where @xmath192 , and @xmath74 are the heat flow and power dropped into the cavity , respectively .
furthermore , we assume that @xmath193 where @xmath194 is the thermal conductivity between the cavity mode volume and the surrounding @xcite . in thermal equilibrium
the heat flow into the cavity will be equal to the heat flow out of the cavity , which allows us to write the equilibrium temperature as @xmath195 next we observe that since @xmath196 where @xmath197 is the fiber taper waveguide input power , we can write @xmath198 and the expected cavity mode shift in terms of the resonance contrast @xmath199\,,\\ & = \lambda_\text{o}\left[1+\left(\frac{a}{k } \frac{\gamma_\text{abs}}{\gamma_\text{tot } } p_{i } \right ) ( 1-\overline{t } ) \right]\,,\\ & = \lambda_\text{o}\left[1+d ( 1-\overline{t } ) \right].\end{aligned}\ ] ] this gives the laser - cavity wavelength detuning , @xmath200 as @xmath201 where @xmath202 is used as a free parameter in fitting our cavity transmission profile .
the laser detuning @xmath86 can then be calculated for any bistable lineshape .
in the weak damping regime ( @xmath203 ) the oscillation amplitude of a thermally driven harmonic oscillator is given by the equipartition theorem @xcite as @xmath204 where @xmath205 is the boltzmann constant , t = 295 k is the bath temperature , and @xmath206 40 pg and @xmath207 ghz are the effective mass and mechanical frequency of the radial breathing mode studied here , respectively .
this results in @xmath208 fm and a zero point fluctuation motion , @xmath209 fm .
while @xmath210 , where @xmath211 is the variance of the mechanical displacement , one must be more careful when calculating the mechanical energy .
strictly speaking @xmath211 is related to the single sided displacement spectral density @xmath212 by @xmath213 this can be connected to the measured cavity transmission noise spectrum @xmath214 through a cavity transfer function @xmath215 , @xmath94 , and @xmath216 @xcite @xmath217 in this experiment we measure @xmath214 , and can compute the area under the curve , @xmath218 , given by @xmath219 . if we change @xmath94 from @xmath220 to @xmath221 , keep @xmath86 constant , and ignore the small ( @xmath222 ) changes in @xmath8 , such that @xmath223 , we can show that the ratio of the area under the curve corresponding to @xmath220 and @xmath221 given by @xmath224 and @xmath225 , respectively , is @xmath226 we can then calibrate high @xmath197 measurements to the thermal case , where @xmath94 is small enough for optomechanical backaction effects to be ignored .
the displacement amplitude , @xmath130 , of the rbm in the self - oscillation regime can then be calculated as @xmath227 where @xmath228 and @xmath229 are the area under the curve in the driven ( @xmath230 ) and thermal ( @xmath231 ) states , respectively .
similarly , for the purpose of comparing mechanical spectra it is useful to calculate the normalized cavity transmission noise spectrum @xmath126 , given by @xmath232 the maximum oscillation amplitude @xmath130 is shown as a function of dropped optical power in fig .
3(a ) , where the absolute maximum oscillation amplitude was found to be @xmath21 31 pm ( @xmath233 ) . using finite element comsol simulations and by assuming that diamond behaves as a linear elastic material in the self oscillation regime , these amplitudes correspond to stress on the order of tens of mpa at the center of the microdisk .
the device studied here demonstrates the largest @xmath66 product of an optomechanical device measured in ambient conditions to date .
figure [ fig : qf ] compares this value with a survey of some of the largest @xmath66 products observed in cavity optomechanical systems in ambient , cryogenic , and low pressure conditions .
note that higher @xmath66 products have been demonstrated compared to this work , but required either vacuum or low - temperature environments . | single - crystal diamond cavity optomechanical devices are a promising example of a hybrid quantum system : by coupling mechanical resonances to both light and electron spins , they can enable new ways for photons to control solid state qubits .
however , realizing cavity optomechanical devices from high quality diamond chips has been an outstanding challenge .
here we demonstrate single - crystal diamond cavity optomechanical devices that can enable photon - phonon - spin coupling . cavity optomechanical coupling to @xmath0 frequency ( @xmath1 ) mechanical resonances is observed . in room
temperature ambient conditions , these resonances have a record combination of low dissipation ( mechanical quality factor , @xmath2 ) and high frequency , with @xmath3 sufficient for room temperature single phonon coherence .
the system exhibits high optical quality factor ( @xmath4 ) resonances at infrared and visible wavelengths , is nearly sideband resolved , and exhibits optomechanical cooperativity @xmath5 .
the devices potential for optomechanical control of diamond electron spins is demonstrated through radiation pressure excitation of mechanical self - oscillations whose 31 pm amplitude is predicted to provide 0.6 mhz coupling rates to diamond nitrogen vacancy center ground state transitions ( 6 hz / phonon ) , and @xmath6 stronger coupling rates to excited state transitions . |
we assume basic familiarity with hopf algebras , as found in the opening chapters of the book @xcite .
let , be a pair of dual hopf algebras over a field @xmath3 .
this means that there is a nondegenerate pairing @xmath4 for which the structure of @xmath5 is dual to that of @xmath6 and vice - versa .
for example , @xmath6 could be finite - dimensional and @xmath5 its linear dual , or @xmath6 could be graded with each component finite - dimensional and @xmath5 its graded dual .
these algebras naturally act on each other @xcite : suppose that @xmath7 and @xmath8 and set @xmath9 ( we use sweedler notation for the coproduct , @xmath10 . )
these left actions are the adjoints of right multiplication : for @xmath11 and @xmath12 , @xmath13 this shows that @xmath5 is a left @xmath6-module under the action in .
in fact , @xmath5 is a left @xmath6module algebra , meaning that for @xmath12 and @xmath7 , @xmath14 recall that the @xmath15 and @xmath16 satisfy @xmath17 and @xmath18 for all @xmath7 .
[ lem : hopf ] for @xmath11 and @xmath8 , we have @xmath19 let @xmath20 .
we prove first the formula @xmath21 ( this is essentially @xmath22 in the proof of lemma 2.1.4 in @xcite . )
expanding the sum using and coassociativity , @xmath23 , gives @xmath24 here , follows as @xmath5 is an @xmath6-module and from the antipode and counit conditions .
note that @xmath25 . using
this becomes @xmath26 which proves the lemma , as this holds for all @xmath20 .
this proof is identical to the argument in the appendix to @xcite , where @xmath27 was a complete homogeneous symmetric function in the hopf algebra @xmath6 of symmetric functions .
we apply lemma [ lem : hopf ] to produce skew littlewood - richardson rules for several hopf algebras in algebraic combinatorics .
we isolate the common features of those arguments . in the notation of section [ sec : harpoon ] , let @xmath28 and @xmath29 be dual bases indexed by some set @xmath30 , so @xmath31 for @xmath32 .
define structure constants for @xmath6 and @xmath5 via @xmath33{0pt}{16pt}\end{aligned}\ ] ] the @xmath34 and @xmath35 defined above co - multiply according to @xmath36 ( note that the structure of @xmath5 can be recovered from the structure of @xmath6 .
thus , we may suppress the analogs of and the second formula in in the coming sections . ) finally , suppose that the antipode acts on @xmath6 in the @xmath37-basis according to the formula @xmath38 for some functions @xmath39 and @xmath40 .
then lemma [ lem : hopf ] takes the following form .
[ th : algebraic rule ] for any @xmath41 , we have @xmath42 swapping @xmath43 and @xmath44 in yields the analog for the skew elements @xmath45 in @xmath5 . the actions in together with the second formulas for the coproducts in and show that @xmath46 and @xmath47 .
now use and to obtain @xmath48 this equals the right hand side of , since @xmath49 .
the condition is highly restrictive .
it implies that the antipode @xmath50 , as a linear map , is conjugate to a signed permutation matrix .
nevertheless , it holds for the hopf algebras we consider .
more generally , it holds if either @xmath6 or @xmath5 is commutative , for then @xmath50 is an involution ( * ? ? ?
the commutative hopf algebra @xmath51 of symmetric functions is graded and self - dual under the hall inner product @xmath52 .
a systematic study of @xmath53 from a hopf algebra perspective appears in @xcite .
we follow the definitions and notation in chapter i of @xcite .
the schur basis of @xmath53 ( indexed by partitions ) is self - dual , so and become @xmath54 where the @xmath55 are the and the @xmath56 are the ( * ? ? ?
combinatorial expressions for the @xmath57 and inner products @xmath58 are derived using the hopf algebraic structure of @xmath53 in @xcite .
the coefficients occur in the triple product @xmath59 , @xmath60 write @xmath61 for the conjugate ( matrix - transpose ) of @xmath62
. then the action of the antipode is @xmath63 which is just a twisted form of the fundamental involution @xmath64 that sends @xmath65 to @xmath66 .
indeed , the formula @xmath67 shows that holds on the generators @xmath68 of @xmath53 .
the validity of follows as both @xmath50 and @xmath64 are algebra maps . since @xmath69 unless @xmath70 ,
we may write as @xmath71 we next formulate a combinatorial version of . given partitions @xmath62 and
@xmath72 , form the skew shape by placing @xmath62 southwest of @xmath72 .
thus , @xmath73 similarly , if @xmath74 is a tableau of shape @xmath62 and @xmath50 a tableau of shape @xmath72 , then @xmath75 is the skew tableau of shape @xmath76 obtained by placing @xmath74 southwest of @xmath50 .
fix a tableau @xmath77 of shape @xmath78 .
the littlewood - richardson coefficient @xmath79 is the number of pairs @xmath80 of tableaux of respective shapes @xmath62 and @xmath72 with @xmath75 knuth - equivalent to @xmath77 .
see ( * ? ? ?
* ch . 5 , cor .
2(v ) ) . similarly ,
@xmath81 is the number of triples @xmath82 of tableaux of respective shapes @xmath83 , @xmath62 , and @xmath72 with @xmath84 knuth - equivalent to @xmath77 .
write for the shape of a tableau @xmath50 and @xmath85 if @xmath50 is knuth - equivalent to @xmath77 .
[ l : delta(skew ) ] let @xmath86 be partitions and fix a tableau @xmath77 of shape @xmath78 .
then @xmath87 the sum taken over triples @xmath88 of tableaux with @xmath89 and @xmath90 .
note that @xmath91 and the operation @xmath92 makes sense for skew tableaux . if @xmath50 is a tableau of skew shape @xmath93 , put @xmath94 .
[ th : two ] let @xmath95 be partitions and fix a tableau @xmath77 of shape @xmath78 .
then @xmath96 the sum taken over triples @xmath97 of skew tableaux of respective shapes @xmath98 , @xmath99 , and @xmath72 such that @xmath100 .
if @xmath77 is the unique of shape @xmath78 whose @xmath101th row contains only the letter @xmath101 , then this is conjecture 6.1 in @xcite .
indeed , in this case @xmath50 is yamanouchi of shape @xmath72 , so the sum is really over pairs of tableaux , and this explains the @xmath72-yamanouchi condition in @xcite .
the difference lies in the tableau @xmath102 and the reading word condition in @xcite .
it is an exercise in tableaux combinatorics that there is a bijection between the indices @xmath103 of theorem [ th : two ] and the corresponding indices of conjecture 6.1 in @xcite .
we reinterpret in terms of tableaux .
let @xmath88 be a triple of tableaux of partition shape with @xmath89 and @xmath104 . if @xmath105 , then by ( * ? ? ?
* ch . 5 , cor .
2(i ) ) , @xmath106 counts skew tableaux @xmath102 of shape @xmath98 with @xmath107 . likewise , if @xmath108 , then @xmath109 counts skew tableaux @xmath110 of shape @xmath99 with @xmath111
. now may be written as @xmath112 summing over skew tableaux @xmath113 with @xmath114 of partition shape , @xmath89 , @xmath90 , @xmath115 , @xmath116 , and @xmath117 .
finally , note that @xmath114 is the unique tableau of partition shape knuth - equivalent to @xmath118 .
since @xmath119 is knuth - equivalent to @xmath77 ( by transitivity of @xmath120 ) , we omit the unnecessary tableaux @xmath114 from the indices of summation and reach the statement of the theorem .
the self - dual hopf algebra of symmetric functions has a natural self - dual subalgebra .
this has dual bases the schur @xmath0- and @xmath1-functions ( * ? ? ?
* iii.8 ) , which are indexed by @xmath121 .
write for the length of the partition @xmath122 . as in section
[ sec : schur s ] , the constants and skew functions in the structure equations @xmath123 have combinatorial interpretations ( see below ) . also , each basis @xmath124 and @xmath125 is _ almost _ self - dual in that @xmath126 and @xmath127 .
the algebra @xmath128 is generated by the special @xmath1-functions @xmath129 ( * ? ? ?
* iii , ( 8.1 ) ) .
this implies that @xmath130 , from which we deduce that @xmath131 as @xmath132 unless @xmath133 , we may write the algebraic rule as @xmath134 with a similar identity holding for @xmath135 ( swapping @xmath136 and @xmath137 ) .
we formulate two combinatorial versions of .
strict partitions @xmath138 are written as ( where row @xmath101 begins in column @xmath101 ) .
skew shifted shapes @xmath139 are defined in the obvious manner : @xmath140 in what follows , tableaux means _ semi - standard ( skew ) shifted tableaux on a marked alphabet _
* iii.8 ) .
we use shifted versions of the jeu - de - taquin and plactic equivalence from @xcite and @xcite , denoting the corresponding relations by and , respectively . given tableaux @xmath141
we write @xmath142 when representative words @xmath143 ( built via `` @xmath144 '' @xcite ) of the corresponding shifted plactic classes satisfy @xmath145
. stembridge notes ( following ( * ? ? ?
8.2 ) ) that for a fixed tableau @xmath146 of shape @xmath147 , @xmath148 serrano has a similar description of these coefficients in terms of @xmath149 .
fixing a tableau @xmath77 of shape @xmath78 , it follows from ( * ? ? ?
1.15 ) that the coefficient @xmath150 in @xmath151 is given by @xmath152 if @xmath77 is a tableau of shape @xmath122 , write @xmath153 for @xmath154 .
the formula relating the @xmath155 s and @xmath156 s combines with and to give our next result .
[ th : schur - q ] let @xmath95 be shifted partitions and fix a tableau @xmath77 of shape @xmath78 . then @xmath157 the sum taken over quintuples @xmath113 with @xmath114 of partition shape , @xmath89 , @xmath158 , @xmath115 , @xmath159 , and @xmath160 .
serrano conjectures an elegant combinatorial description ( * ? ? ?
* conj . 2.12 and cor . 2.13 ) of the structure constants @xmath161 in : for any tableau @xmath146 of shape @xmath147 , he conjectured @xmath162 ( note that if @xmath163 are tableaux , then @xmath164 does not necessarily imply that @xmath165 . )
this leads to a conjectural reformulation of theorem [ th : schur - q ] in the spirit of theorem [ th : two ] .
[ th : conj schur - q ] let @xmath95 be partitions and fix a tableau @xmath77 of shape @xmath78 . then @xmath166 the sum taken over triples @xmath97 of skew tableaux of respective shapes
@xmath167 , @xmath99 , and @xmath72 such that @xmath168
. there is a unique shifted tableau @xmath74 in any shifted plactic class ( * ? ?
so the conditions @xmath169 and @xmath170 in and may be replaced with the single condition @xmath168 .
the hopf algebra of noncommutative symmetric functions was introduced , independently , in @xcite as the ( graded ) dual to the commutative hopf algebra of quasisymmetric functions .
we consider the dual bases ( indexed by compositions ) @xmath171 of gessel s quasisymmetric functions and @xmath172 of noncommutative ribbon schur functions .
the structure constants in @xmath173 may be given combinatorial meaning via the descent map @xmath174 from permutations to compositions and a section of it @xmath175 .
these maps are linked via , edge - connected skew young diagrams ( written in the french style ) , with no @xmath176 subdiagram present . by way of example ,
@xmath177 ( in the intermediate step for @xmath178 , new rows in the ribbon begin at descents of @xmath179 . in the intermediate step for @xmath180
, the boxes in the ribbon are filled left - to - right , bottom - to - top . )
a ribbon @xmath181 may be extended by a ribbon @xmath182 in two ways : affixing @xmath182 to the rightmost edge or bottommost edge of @xmath181 ( written @xmath183 and @xmath184 , respectively ) : @xmath185 if a ribbon @xmath186 is formed from @xmath181 and @xmath182 in either of these two ways , we write @xmath187 .
the coefficient @xmath188 is @xmath189 , if @xmath187 , and @xmath190 otherwise .
if @xmath191 is the shifted shuffle product on permutations ( see ( 3.4 ) in @xcite ) , then the coefficient @xmath192 is the number of words @xmath179 in @xmath193 such that @xmath194 .
the coefficient @xmath195 has the analogous description .
antipode formulas for the distinquished bases were found , independently , in @xcite : @xmath196 where @xmath197 is the conjugate of @xmath181 ( in the sense of french style skew partitions ) .
for example , @xmath198 .
the descriptions of the antipode and structure constants in and give a formula for the product of two skew ribbon schur functions .
[ th : schur - r ] let @xmath199 be compositions
. then @xmath200 the sum taken over factorizations @xmath201 , extensions @xmath202 , and words @xmath179 in the shuffle product @xmath203 such that @xmath204 .
[ rem : f skew ] the nonzero skew ribbon schur functions @xmath205 do not correspond to skew ribbon shapes in a simple way .
for example , @xmath206 is not a ( connected ) sub - ribbon of @xmath207 , yet @xmath208 .
contrast this with the skew functions @xmath209 , where @xmath210 if and only if @xmath211 for some ribbon @xmath64 .
that is , @xmath212 .
thus we may view @xmath181 as the last @xmath213 boxes of the ribbon @xmath182 and @xmath214 as the complementary ribbon @xmath64 .
interpreting @xmath215 alternately as a product of ordinary functions or skew functions yields the curious identity @xmath216
fix an integer @xmath217 . let @xmath218 denote the hopf subalgebra of the hopf algebra of symmetric functions generated by the homogeneous symmetric functions @xmath219 .
let @xmath220 denote the hopf - dual quotient hopf algebra of symmetric functions .
we consider the dual bases @xmath221 and @xmath222 of @xmath2-schur functions and dual @xmath2-schur functions of @xcite , also called strong schur functions and weak schur functions in @xcite .
the @xmath2-schur functions were first introduced by lapointe , lascoux , and morse in the context of macdonald polynomials , and were later shown by lam to represent schubert classes in the affine grassmannian of @xmath223 .
we refer the reader to the references in @xcite .
here @xmath122 varies over all @xmath2-bounded partitions , that is , those partitions satisfying @xmath224 .
there is an involution @xmath225 on @xmath2-bounded partitions called .
we have @xmath226 if @xmath227 is a one - part partition , then @xmath228 is a homogeneous symmetric function .
we have the @xmath2-pieri and dual @xmath2-pieri rules @xcite ( called weak and strong pieri rules in @xcite ) @xmath229 for @xmath230 . here
@xmath231 denotes an connecting @xmath122 and @xmath147present if and only if both @xmath93 and @xmath232 are horizontal @xmath233-strips .
the notation @xmath234 denotes an as introduced in @xcite , which we will not define here .
the terminology comes from the relationship with the weak and strong ( bruhat ) orders of the affine symmetric group .
we remark that there may be distinct strong strips @xmath234 and @xmath235 which start and end at the same partition , so that the second pieri rule of may have multiplicities .
( strictly speaking , the strong strips in @xcite are built on @xmath236-cores , and our @xmath234 denotes the strips obtained after applying a bijection between @xmath236-cores and @xmath2-bounded partitions . )
we define skew functions @xmath237 and @xmath238 using and .
there is an explicit combinatorial description of @xmath238 in terms of the weak tableaux of @xcite , but only a conjectured combinatorial description of @xmath237 ( * ? ? ?
4.18(3 ) ) .
[ thm : skewkpieri ] for @xmath2-bounded partitions @xmath239 , and @xmath230 , @xmath240 ( \lambda^-)^{\omega_k } { { \,\rightsquigarrow_{j}\ , } } \lambda^{\omega_k } } } s_{\mu^+/\lambda^-}^{(k)}\ ] ] in theorem [ th : algebraic rule ] , take @xmath241 . for @xmath81 , use the formula @xmath242 , and for @xmath243 and @xmath244 , use .
[ thm : skewdualkpieri ] for @xmath2-bounded partitions @xmath245 , and @xmath230 , @xmath246 ( \mu^-)^{\omega_k } { { \,\rightarrow_{j}\ , } } \mu^{\omega_k } } } f_{\lambda^+/\mu^-}^{(k)}\ ] ] identical to the proof of theorem [ thm : skewkpieri ] . as an example , let @xmath247 , @xmath248 , @xmath249 , and @xmath250 .
then theorem [ thm : skewkpieri ] states that @xmath251 which one can verify using and the expansions @xmath252 and @xmath253 .
theorem [ thm : skewdualkpieri ] states that @xmath254 one can verify that both sides are equal to @xmath255 .
fix @xmath256 . the hopf algebra @xmath128 of section [ sec : schur pq ] contains a hopf subalgebra @xmath257 generated by the schur @xmath0-functions @xmath258 . in @xcite
, it was shown that @xmath257 is isomorphic to the homology ring @xmath259 of the affine grassmannian of the symplectic group @xmath260 .
a distinguished basis @xmath261 , representing the schubert basis , was studied there .
the symmetric functions @xmath262 are shifted versions of the @xmath2-schur functions of section [ sec : kschur ] .
the indexing set for the basis @xmath263 is the set @xmath264 of affine grassmannian type @xmath265 permutations : they are the minimal length coset representatives of @xmath266 in @xmath267 .
a lower bruhat order ideal @xmath268 of the affine type @xmath265 weyl group is defined in @xcite .
let @xmath269 denote those @xmath270 with length @xmath271 . for each @xmath272
, there is a nonnegative integer @xmath273 , called the of @xmath274 .
we note that @xmath275 . with this notation , for each @xmath276 , we have the pieri rule ( * ? ? ?
1.3 and 1.4 ) @xmath277 where the sum is over all @xmath278 such that @xmath279 , and @xmath280 . in theorem [ th : algebraic rule ] , take @xmath292 and use . for
the constants @xmath195 , use the formula @xmath293 if @xmath294 , the product @xmath295 in becomes @xmath296 . if @xmath297 ( resp . , @xmath298 )
, it becomes @xmath299 ( resp . , @xmath300 ) .
susan montgomery , _ hopf algebras and their actions on rings _ , cbms regional conference series in mathematics , vol . 82 , published for the conference board of the mathematical sciences , washington , dc , 1993 . | we use hopf algebras to prove a version of the littlewood - richardson rule for skew schur functions , which implies a conjecture of assaf and mcnamara .
we also establish skew littlewood - richardson rules for schur @xmath0- and @xmath1-functions and noncommutative ribbon schur functions , as well as skew pieri rules for @xmath2-schur functions , dual @xmath2-schur functions , and for the homology of the affine grassmannian of the symplectic group .
assaf and mcnamara @xcite recently used combinatorics to give an elegant and surprising formula for the product of a skew schur function and a complete homogeneous symmetric function .
their paper included a conjectural skew version of the littlewood - richardson rule , and also an appendix by one of us ( lam ) with a simple algebraic proof of their formula .
we show how these formulas and much more are special cases of a simple formula that holds for any pair of dual hopf algebras .
we first establish this hopf - algebraic formula , and then apply it to obtain formulas in some well - known hopf algebras in combinatorics . |
in classical physics one has time symmetric dynamical equations with time asymmetric boundary conditions @xcite .
these time asymmetric boundary conditions come in pairs : given one time asymmetric boundary condition , its time reversed boundary condition can also be formulated mathematically .
for example in classical electrodynamics one has retarded and advanced solutions of the time symmetric dynamical ( maxwell ) equations or in general relativity one has time asymmetric big bang and big crunch solutions of einstein s time symmetric equation . except for a few prominent cases of pedagogical importance ( e.g. stationary states or cyclic evolutions ) ,
the physics of our world is predominantly time - asymmetric .
somehow nature chooses one of the pair of time asymmetric boundary conditions .
the standard quantum mechanics in hilbert space @xcite does not allow time asymmetric boundary conditions for the schrdinger or von neumann equation @xcite .
however this is a consequence of the _ mathematical _ properties of the hilbert space and need not imply that quantum _ physics _ is strictly time symmetric
. it would be incredible if classical electrodynamics had a radiation arrow of time and quantum electrodynamics did not also have an arrow of time
. in quantum physics peierls and siegert considered many years ago time asymmetric solutions with purely outgoing boundary conditions @xcite .
the choice of appropriate dense subspaces @xmath2 and @xmath3 of the ( complete ) hilbert space @xmath4 allows the formulation of time asymmetric boundary conditions : @xmath5{3.25in}{for the out - states $ \{\psi^-\}$ of scattering theory which are actually observables as defined by the registration apparatus ( detector ) , and}\nonumber\\ \phi_- \subset { \ensuremath{\mathcal{h } } } & & \parbox[t]{3.25in}{for the in - states $ \{\phi^+\}$ which are prepared states as defined by the preparation apparatus ( accelerator ) . } \nonumber\end{aligned}\ ] ] time asymmetric quantum theory distinguishes meticulously between states @xmath6 and observables @xmath7 .
two different dense subspaces of the hilbert space @xmath4 are chosen , @xmath8 and @xmath9 .
the standard hilbert space quantum theory uses @xmath10 for both , @xmath11 and as a result is time symmetric with a reversible unitary group time evolution . in the theory of scattering and decay , a pair of time asymmetric boundary conditions can be heuristically implemented by choosing in- and out - plane wave `` states '' @xmath12 and @xmath13 which are solutions of the lippmann - schwinger equation @xcite @xmath14 where @xmath15 .
the energy distribution of the incident beam is given by @xmath16 and the energy resolution of the detector ( which for perfect efficiency is the energy distribution of the detected `` out - states '' ) is measured as @xmath17 are the basis system that is used for the dirac basis vector expansion of the in - states @xmath18 and the out - states ( observables ) @xmath19 : @xmath20 where @xmath21 are the degeneracy labels .
the dirac kets of the lippmann - schwinger equation @xmath22 are in our time asymmetric quantum theory @xcite antilinear functionals on the spaces @xmath23 , i.e. they are elements of the dual space @xmath24 .
this leads to two rigged hilbert spaces ( rhs ) and the following new hypothesis for time asymmetric quantum theory and general observables by ( positive definite ) operators @xmath25 . ] : @xmath26{3in}{pure registered observables or so - called `` out - states '' are described by the vectors } & & \psi^- \in \phi_+\subset{\ensuremath{\mathcal{h}}}\subset{\ensuremath{\phi^\times}}_+\nonumber\\ \parbox[c]{3in}{and pure prepared in - states are described by the vectors}&&\phi^+ \in \phi_-\subset{\ensuremath{\mathcal{h}}}\subset{\ensuremath{\phi^\times}}_-.\label{rhs 's } \end{aligned}\ ] ] this new hypothesis with the appropriate choice for the spaces @xmath2 and @xmath3 given below in ( [ caus])is essentially all by which our time asymmetric quantum theory differs from the standard hilbert space quantum mechanics which imposes @xmath11 ( or @xmath27 ) .
in addition to the dirac lippmann - schwinger kets @xmath24 , the dual spaces @xmath28 of the rhs s also contain gamow kets @xmath29 , which are generalized eigenvectors of the ( self - adjoint ) hamiltonian with complex eigenvalue @xmath30 .
we use these gamow kets to describe quasistable particles .
we shall now mathematically define @xmath2 and @xmath3 , and therewith the rhs s ( [ rhs s ] ) . from a mathematical formulation of causality expressed by the truism ``
a state must be prepared before an observable can be measured ( registered ) in it '' , one can argue that the energy wave functions @xmath31 are the boundary values of analytic functions in the upper half energy plane ( second sheet of the @xmath0-matrix ) and the @xmath32 are the same for the lower half plane @xcite .
precisely , @xmath33 where @xmath34 is the schwartz space and @xmath35 are well - behaved hardy functions in the lower ( upper ) half plane @xmath36 of the second riemann sheet for the @xmath0-matrix @xmath37 .
the disparity between the labels @xmath38 for the vectors and the spaces ( e.g. @xmath18 ) now makes sense .
the superscripts of the vectors is the standard notation of scattering theory while the subscripts of the spaces comes from their mathematical definition ( [ caus ] ) .
this correspondence ( [ caus ] ) between the physical state vectors and the mathematical spaces is a wonderful example of what wigner calls `` the unreasonalbe effectiveness of mathematics in the natural sciences '' @xcite .
stable states are described by bound state poles or by eigenvectors of the self - adjoint hamiltonian @xmath39 with real eigenvalue @xmath40 : @xmath41 quantum mechanical resonances are most commonly defined by the pair of resonance poles in the second riemann sheet of the analytically continued @xmath0-matrix at the position @xmath42 ; a decaying state is associated to the pole at @xmath43 .
vector representations of quasistable particles like the @xmath44 mesons in the lee - oehme - yang theory @xcite use approximate methods such as the weisskopf - wigner approximation @xcite for which `` there does not exist ... a rigorous theory to which these various methods can be considered approximations '' @xcite .
often one hears the opinion that resonances and decaying states are complicated objects and can not be represented by simple ( exponentially decaying ) state vectors .
this opinion differs from the common practice to classify quasistable particles along with stable particles @xcite and also contradicts empirical facts : quasistable particles are not qualitatively different from stable particles but only quantitatively by a non - zero value of the width @xmath45 which one takes to be the inverse lifetime , @xmath46 .
stability or the value of the lifetime is not a criterion for elementarity .
a particle decays if it can and remains stable if selection rules for some quantum numbers prevents it from decaying .
stable and quasistable states should be described on the same footing , e.g. both defined by a pole of the @xmath0-matrix at the position @xmath40 for stable particles and at the position @xmath43 for quasistable particles . then in analogy to the vector description ( [ stable ] ) for stable particles , there should also be a description of a quasistable particle by a generalized eigenvector with eigenvalue @xmath47 . a vector description is also needed to express the initial decay rate by the golden rule as in @xcite .
therefore , in analogy to ( [ stable ] ) resonances and quasistable particles will be described as generalized eigenvectors of the self - adjoint hamiltonian with complex eigenvalue @xmath48 , where @xmath49 represents the resonance energy and @xmath50 is the lifetime : @xmath51 where the only distinction from ( [ stable ] ) is that @xmath52 are bra - kets ( @xmath2-continuous functionals ) and @xmath53 are scalar products ( @xmath10-continuous functionals ) .
we call the vectors @xmath54 gamow vectors . to obtain this vector representation of quasistable states
one starts with the @xmath0-matrix element @xmath55 and analytically continues the integral to the resonance pole @xmath56 of the @xmath0-matrix @xmath57 @xcite as shown in fig . 1 .
for this to be possible , the set of in - states @xmath58 and out - states @xmath59 must have certain analyticity properties . in order to get a breit - wigner energy distribution for the pole term of ( [ smat ] ) , the energy wave functions @xmath60 and @xmath61 must be well - behaved hardy class functions of the upper and lower half plane , respectively , in the second sheet of the energy surface of the @xmath0-matrix .
this is the same mathematical condition as provided by causality and stated by ( [ caus ] ) . using the cauchy formula
, the analytically continued kets @xmath62 become the gamow kets @xmath63 at the resonance pole @xmath47 .
for the gamow kets one can prove the equations ( [ unstable ] ) the first equality of ( [ unstable ] ) is the definition of the conjugate operator @xmath64 in @xmath65 , which is a unique extension of the hamiltonian @xmath66 in @xmath4 .
the second equality of ( [ unstable ] ) says that @xmath67 is a generalized eigenvector of @xmath39 with a complex eigenvalue , similar to @xmath68 , which in dirac s notation is written @xmath69 .
these gamow kets @xmath67 have all the properties required of a vector representing the `` state '' of an unstable particle or ( together with the ket @xmath70 for the @xmath0-matrix pole at @xmath71 ) of a resonance in non - relativistic quantum physics .
gamow vectors have the following features : 1 .
they are derived as functionals of the resonance pole term at @xmath43 ( and at @xmath72 ) in the second sheet of the analytically continued @xmath0-matrix .
the gamow vectors have a breit - wigner energy distribution : @xmath73 3 .
they have an asymmetric time evolution and obey an exponential law : @xmath74 4 .
the decay probability of @xmath75 into the final non - interacting decay products described by @xmath76 , @xmath77 , can be calculated as a function of time for @xmath78 , and from this the decay rate @xmath79 is obtained by differentiation .
this leads to an exact golden rule ( with the natural lineshape given by a breit - wigner ) which in the born approximation ( @xmath80 ; @xmath81 ; @xmath82 , where @xmath83 is an eigenvector of @xmath84 with eigenvalue @xmath85 ) goes into fermi s golden rule no . 2 of dirac .
stable relativistic particles are defined and described by unitary irreducible representations @xmath86 of the poincar group @xmath87 @xcite where @xmath88 is the eigenvalue of @xmath89 and @xmath90 is the spin . in the irreducible representation space
one commonly uses the wigner basis vectors @xmath91 , but one could just as well use any other complete basis system , e.g. the 4-velocity basis @xmath92 where @xmath93 and @xmath94 with @xmath95 this means every vector @xmath96 can be written as ( dirac basis vector expansion ) : @xmath97 stable relativistic particles are characterized in addition to the charge quantum numbers or particle labels by the value of mass @xmath98 and spin @xmath90 .
they are therefore described by the vectors @xmath99 or @xmath100 of the unitary irreducible representation spaces ( [ vector ] ) .
they lead to poles of the @xmath90th partial @xmath0-matrix @xmath101 at the value @xmath102 on the real axis of the first sheet @xcite .
decaying states or resonances are defined by poles of the @xmath90th partial @xmath0-matrix @xmath101 on the second sheet at @xmath103 where @xmath104 and @xmath105 are different parameterizations of the complex value @xmath106 . here
@xmath107 is the invariant mass squared , i.e. the eigenvalue of the total mass operator @xmath108 , @xmath109 .
the spin @xmath90 is the total angular momentum of the @xmath110 decay products and is equal to the resonance spin @xmath111 .
quasistable relativistic particles are thus characterized by @xmath112 and by spin @xmath113 , where we call the resonance mass @xmath114 and the resonance width @xmath115 for reason seen below in ( [ evol ] ) .
more common is to call @xmath116 and @xmath117 of ( [ para ] ) the resonance mass and width @xcite .
some examples are : 1 .
relativistic hadron resonances such as the @xmath118 meson in @xmath119 with @xmath120 .
2 . relativistic ( weakly ) decaying states such as @xmath121 with @xmath122 .
3 . the @xmath1-bosons in @xmath123 with @xmath124 . for these
resonance scattering and decay processes one uses in the @xmath0-matrix elements ( [ smat ] ) , the relativistic dirac lippmann - schwinger `` scattering states '' @xmath125 for `` physical '' values of @xmath88 , or in other words for values of @xmath88 with @xmath126 . to obtain
the relativistic gamow kets one analytically continues these relativistic dirac lippmann - schwinger kets in the contour deformation for the integral of ( [ smat ] ) in the following way : the real @xmath127 becomes a complex @xmath88 but the @xmath128 _ remain real_. at the pole position @xmath129 one obtains the relativistic gamow kets : @xmath130 these gamow kets are basis vectors of a `` minimally complex '' semigroup representation of @xmath87 which is characterized by @xmath131 .
this is the analogy of the representation for non - relativistic gamow vectors @xcite : @xmath132 the relativistic gamow vectors have according to ( [ rgk ] ) a `` relativistic breit - wigner '' @xmath88-distribution given by @xmath133 .
they can be shown to be generalized eigenvectors of the invariant mass - squared operator @xmath134 @xcite , i.e. they fulfill @xmath135 for every @xmath136 .
the lorentz transformations @xmath76 in these minimally complex representations are represented by a group of unitary operators @xmath137 and act in the well - known way : @xmath138 however the space - time translations can no more be represented by a group of operators @xmath139 in @xmath140 the rigged hilbert spaces of hardy class have to be employed in rigorously obtaining ( [ rgk ] ) and ( [ eq : action.of.m2 ] ) and then one can see that only the semigroup of space - time translations into the forward light cone can be represented in the space ( [ phiplus ] ) .
much like in the rhs theory of non - relativistic gamow vectors , one can show that the time translation of the decaying state in the rest frame @xmath141 is given by @xmath142 where @xmath143 is the time in the rest system .
thus , the relativistic gamow vectors have a semigroup time evolution and obey the exponential law . from ( [ evol ] )
it follows that the lifetime of the particle represented by the relativistic gamow vector is given by @xmath144 where the width @xmath115 is the @xmath145 given by the resonance pole position @xmath106 of the relativistic @xmath0-matrix .
the real resonance mass @xmath114 in ( [ evol ] ) is the @xmath146 and therefore not exactly the same as the peak position @xmath147 of the modulus of the relativistic breit - wigner amplitude of ( [ rgk ] ) .
our fundamental assertion is that stable and unstable particles are not fundamentally different .
the theoretical treatment of stable and unstable particles should therefore be the same . within the hilbert space ( hs ) quantum theory , time asymmetric boundary conditions can not be formulated and a theory of decaying states within the hs framework can only be approximate @xcite . within the rhs , time asymmetric quantum theory can be introduced by the new hypothesis ( [ rhss],[caus ] ) which allows a distinction between states and observables . within this theory , gamow vectors
, both relativistic and non - relativistic , satisfy all necessary properties of representing quasi - stable states and resonances .
they are associated to poles of the @xmath0-matrix , have breit - wigner energy distribution and obey an exponential semigroup time evolution .
all of these properties can be formulated mathematically precisely , and a particular property of the gamow vector is that it relates the width @xmath115 of the breit - wigner energy distribution and the lifetime @xmath148 of the exponential decay @xmath149 . without the gamow vectors ,
the width and inverse lifetime only can be shown to be approximately equal by use of some approximate method based on weisskopf - wigner @xcite .
consequently , one did not know whether to choose @xmath150 or @xmath105 of ( [ para ] ) or any other @xmath151 as the mass and width of relativistic resonances .
the relativistic gamow vectors unify in a fundamental picture both stable and unstable relativistic particles .
both are given by irreducible representations of poincar transformations .
stable particles are unitary representations characterized by a real mass and have unitary group time evolution , quasi - stable particles are `` minimally complex '' semigroup representations characterized by a complex mass and have semigroup time evolution with an arrow of time .
that relativistic gamow states @xmath67 possess only semigroup transformations into the forward light cone ( and that others possess only those into the backward light cone ) and no space - like translations ( for either case ) is a result whose interpretation is not yet clear .
it is the relativistic analogue of the semigroup time evolution in non - relativistic quantum mechanics which can be understood as the causality condition that a state must be first prepared before an observable can be measured in it @xcite .
this time asymmetry on the microphysical level
irreversibility without entropy increase and without violation of time reversal invariance was the most surprising and remarkable property of the non - relativistic and relativistic gamow vectors .
these lecture notes are dedicated to professor h. d. doebner on the occasion of his graduation to emeritus status .
we gratefully acknowledge valuable support from the welch foundation .
p. l. kapur and r. e. peierls , proc .
soc . a * 166 * , 277 ( 1938 ) ; a. f. j. siegert , phys . rev . * 56 * , 750 ( 1939 ) ; g. garcia - calderon and r. peierls , nucl .
a * 256 * , 443 ( 1976 ) ; e. hernandez and a. mondragon , phys .
c * 29 * , 722 ( 1984 ) ; a. mondragon and e. hernandez , ann .
der physik * 48 * , 503 ( 1991 ) ; g. garcia - calderon , _ symmetries in physics _
( moshinsky symposium ) , edited by a. frank and k. b. wolf , ( springer - verlag , berlin , 1992 ) .
v. weisskopf and e. p. wigner , z. physik * 63 * , 54 ( 1930 ) ; _ ibid . _ * 65 * , 18 ( 1930 ) ; w. heitler , _ quantum theory of radiation _
( oxford , new york , 1954 ) ; m. l. goldberger and k. m. watson , _ collision theory _ ( wiley , new york , 1964 ) ; r. g. newton , _ scattering theory of waves and particles _ , 2nd ed .
( springer - verlag , berlin , 1982 ) .
e. p. wigner in _ group theoretical concepts in elementary particle physics _ , f. gursey , ed .
, p. 37 - 80 , ( gordan and breach , new york , 1964 ) ; s. weinberg , _ the quantum theory of fields _ ,
vol.1 , ( cambridge university press , 1995 ) ; h. joos , fortsch .
physik * 10 * , 65 ( 1962 ) .
a. bohm , h. kaldass , s. wickramasekara and p. kielanowski , `` semigroup representations of the poincar group and relativistic gamow vectors , '' lanl archives ( hep - th/9911059 ) ; physics letters a , to appear . | the foundations of time asymmetric quantum theory are reviewed and are applied to the construction of relativistic gamow vectors .
relativistic gamow vectors are obtained from the resonance pole of the @xmath0-matrix and furnish an irreducible representation of the poincar semigroup .
they have all the properties needed to represent relativistic quasistable particles and can be used to fix the definition of mass and width of relativistic resonances like the @xmath1-boson .
most remarkably , they have only a semigroup time evolution into the forward light cone expressing time asymmetry on the microphysical level . 1.5 cm 0.5 cm physics department + the university of texas at austin + austin , texas 78712 , usa |
in linear algebra it is frequently necessary to use non - linear objects such as minimal and characteristic polynomials since they encode fundamental information about endomorphisms of finite - dimensional vector spaces .
it is well - known that if @xmath2 is a field and @xmath3 is a zero - dimensional affine @xmath2-algebra , _
i.e. _ a zero - dimensional algebra of type @xmath4/i$ ] , then @xmath3 is a finite - dimensional @xmath2-vector space ( see proposition 3.7.1 of @xcite ) .
consequently , it not surprising that minimal and characteristic polynomials can be successfully used to detect properties of @xmath3 .
this point of view was taken systematically in the new book @xcite where the particular importance of minimal polynomials ( rather greater than that of characteristic polynomials ) emerged quite clearly .
the book also described several algorithms which use minimal polynomials as a crucial tool .
the approach taken there was a good source of inspiration for our research , so that we decided to dig into the theory of minimal polynomials , their uses , and their applications .
the first step was to implement algorithms for computing the minimal polynomial of an element of @xmath3 and of a @xmath2-endomorphism of @xmath3 ( see algorithms [ alg : minpolyquotelim ] , [ alg : minpolyquotmat ] , [ alg : minpolyquotdef ] ) .
they are described in section 3 , refine similar algorithms examined in @xcite , and have been implemented in ( see @xcite and @xcite ) as indeed have all other algorithms described in this paper .
sections [ reductions of ideals ] and [ modularapproach ] constitute a contribution of decisive practical significance : they address the problem of computing minimal polynomials of elements of a @xmath5-algebra using a modular approach . as always with a modular approach ,
various obstacles have to be overcome ( see for instance the discussion contained in @xcite ) .
in particular , we deal with the notion of reduction of an ideal modulo @xmath6 , and we do it by introducing the notion of @xmath7-denominator of an ideal ( see definition [ reductionmodp ] and theorem [ thm : rgbp ] ) , which enables us to overcome the obstacles . then ugly , usable , good and bad primes show up ( see definition [ def : usableprime ] and [ def : goodprime ] ) . fortunately , almost all primes are good ( see theorem [ almostallgood ] and corollary [ cor : badprimes ] ) which paves the way to the construction of the fundamental algorithm [ alg : modular ] .
section 6 presents non - trivial examples of minimal polynomials computed with , and section 7 shows how minimal polynomials can be successfully and efficiently used to compute several important invariants of zero - dimensional affine @xmath2-algebras .
more specifically , in subsection [ isradical and radical ] we describe algorithms [ alg : isradical0dim ] and [ alg : radical0dim ] which show how to determine whether a zero - dimensional ideal is radical , and how to compute the radical of a zero - dimensional ideal . in subsection [ ismaximal ] we present several algorithms
which determine whether a zero - dimensional ideal is maximal or primary .
the techniques used depend very much on the field @xmath2 .
the main distinction is between small finite fields and fields of characteristic zero or big fields of positive characteristic . in particular , it is noteworthy that in the first case frobenius spaces ( see section 5.2 of @xcite ) play a fundamental role . finally , in section [ primary decomposition ]
a series of algorithms ( see [ alg : pdsplitting ] , [ alg : pdsplittingchar0 ] , [ alg : primarydecompositioncore ] , and [ alg : primarydecomposition0dim ] ) describe how to compute the primary decomposition of a zero - dimensional affine @xmath2-algebra .
they are inspired by the content of chapter 5 of @xcite , but they present many novelties . as we said , all the algorithms described in this paper have been implemented in .
their merits are also illustrated by the tables of examples contained in sections [ timings ] and at the end of section [ uses ] .
the experiments were performed on a macbook pro 2.9ghz intel core i7 , using our implementation in 5 .
here we introduce the notation and terminology we shall use and the definition of minimal polynomial which is the fundamental object studied in the paper .
let @xmath2 be a field , let @xmath8 $ ] be a polynomial ring in @xmath9 indeterminates , and let @xmath10 denote the monoid of power products in @xmath11 .
let @xmath0 be a zero - dimensional ideal in @xmath12 ; this implies that the ring @xmath13 is a zero - dimensional affine @xmath2-algebra , hence it is a finite dimensional @xmath2-vector space . then , for any @xmath14 in @xmath12 there is a linear dependency mod @xmath0 among the powers of @xmath14 : in other words , there is a polynomial @xmath15 $ ] which vanishes modulo @xmath0 when evaluated at @xmath16 .
[ def : minpoly ] let @xmath2 be a field , let @xmath17 $ ] , and let @xmath0 be a zero - dimensional ideal .
given a polynomial @xmath18 , we have a @xmath2-algebra homomorphism @xmath19 \to p / i$ ] given by @xmath20 .
the monic generator of the kernel of this homomorphism is called the * minimal polynomial * of @xmath21 ( or simply `` of @xmath14 '' when the ideal @xmath0 is obvious ) , and is denoted by @xmath22 .
the particular case of @xmath23 , where @xmath24 is an indeterminate , is a very important and popular object when computing : in fact @xmath23 is the lowest degree polynomial in @xmath24 belonging to @xmath0 , that is @xmath25 = { \langle { { \mu_{x_i , i}}}(x_i ) \rangle}$ ] .
it is well known that this polynomial may be computed via elimination of all the other indeterminates @xmath26 ( see for example corollary 3.4.6 of @xcite ) .
however the algorithm which derives from this observation is usually impractically slow .
[ rem : nfsigma ] for the basic properties of grbner bases we refer to @xcite .
let @xmath7 be a term ordering on @xmath10 , and let @xmath0 be an ideal in the polynomial ring @xmath12 .
for every polynomial @xmath27 it is known that @xmath28 , the @xmath7-normal form of @xmath14 with respect to @xmath0 , does not depend on which @xmath7-grbner basis of @xmath0 is used nor on which specific rewriting steps were used to calculate it ( see proposition 2.4.7 of @xcite ) . if @xmath0 is clear from the context , we write simply @xmath29 . [ rem : zdelta ] following convention , for @xmath30 we use the symbol @xmath31 to denote the * localization * of @xmath32 by the multiplicative set generated by @xmath33 , _
i.e. _ the subring of @xmath5 consisting of numbers represented by fractions of type @xmath34 where @xmath35 , @xmath36
. observe that @xmath31 depends only on the radical @xmath37 , _ i.e. _ the product of all primes dividing @xmath33 .
furthermore , if @xmath38 then @xmath39 divides @xmath40 if and only if @xmath41 is a subring of @xmath42 .
if @xmath6 is a prime number we use the symbol @xmath43 to denote the * finite field * @xmath44 .
note that some authors write @xmath45 to mean the field @xmath43 : this is clearly ambiguous .
let @xmath2 be a field , let @xmath17 $ ] , let @xmath0 be a zero - dimensional ideal , and let @xmath18 .
a well - known method for computing @xmath22 is by elimination .
one extends @xmath12 with a new indeterminate to produce @xmath46 $ ] , then defines the ideal @xmath47 in @xmath3 , and finally eliminates the indeterminates @xmath48 .
here we give a refined version of algorithm 5.1.1 of @xcite . *
( minpolyquotelim ) * [ alg : minpolyquotelim ] _ notation : _ : : @xmath49 $ ] input : : @xmath0 , a zero - dimensional ideal in @xmath12 , and a polynomial @xmath50 1 : : create the polynomial ring @xmath51 $ ] 2 : : define the ideal @xmath52 3 : : return * the monic , minimal generator of @xmath53 $ ] * output : : @xmath54 $ ] another way to compute @xmath22 is via multiplication endomorphisms on @xmath1 .
let @xmath18 then we write @xmath55 for the endomorphism `` multiplication by @xmath56 '' .
there is a natural isomorphism between @xmath1 and @xmath57 $ ] associating @xmath56 with @xmath58 ( see proposition 4.1.2 in @xcite ) .
the minimal polynomial of @xmath14 with respect to @xmath0 is the same as the minimal polynomial of the endomorphism @xmath58 .
thus , if the matrix @xmath59 represents @xmath58 with respect to some @xmath2-basis of @xmath1 , we can compute the minimal polynomial of @xmath59 ( and thus of @xmath58 ) using the following algorithm which is a refined version of algorithm 1.1.8 of @xcite .
* ( minpolyquotmat ) * [ alg : minpolyquotmat ] _ notation : _ : : @xmath49 $ ] with term ordering @xmath7 input : : @xmath0 , a zero - dimensional ideal in @xmath12 , and a polynomial @xmath50 1 : : compute @xmath60 , a @xmath7-grbner basis for @xmath0 ; + from @xmath60 compute @xmath61 , the corresponding monomial quotient basis of @xmath1 2 : : compute @xmath59 , the matrix representing the map @xmath58 w.r.t .
@xmath61 3 : : let @xmath62 and @xmath63 4 : : _ main loop : _ for @xmath64 do + 4.1 ; ; let @xmath65 ( hence @xmath66 ) 4.2 ; ; is there a linear dependency @xmath67 , @xmath68 ?
+ yes : : return * @xmath69 * no : : append @xmath70 to @xmath71 output : : @xmath54 $ ] note that in step minpolyquotmat-4.2 a linear dependency among the matrices @xmath72 is equivalent to a linear dependency among just their _ first columns_. the reason is that the first column contains the coefficients of @xmath73 assuming that the monomial @xmath74 appears as the first element of @xmath61 .
there is a still more direct approach .
it comes from considering the very definition of minimal polynomial : we look for the first linear dependency among the powers @xmath75 in @xmath1 .
here we give a refined version of algorithm 5.1.2 of @xcite . *
( minpolyquotdef ) * [ alg : minpolyquotdef ] _ notation : _ : : @xmath49 $ ] with term ordering @xmath7 input : : @xmath0 , a zero - dimensional ideal in @xmath12 , and a polynomial @xmath50 1 : : compute @xmath60 , a @xmath7-grbner basis for @xmath0 ; + from @xmath60 compute @xmath61 , the corresponding monomial quotient basis of @xmath1 2 : : let @xmath76 3 : : let @xmath77 and @xmath78 4 : : _ main loop : _ for @xmath64 do + 4.1 ; ; compute @xmath79 ; consequently @xmath80 4.2 ; ; is there a linear dependency @xmath81 , @xmath68
? + yes : : return * @xmath69 * no : : append @xmath82 to @xmath71 output : : @xmath54 $ ] notice that minpolyquotmat and minpolyquotdef essentially do the same computation : the first using a matrix ( dense ) representation , and the second a polynomial ( sparse ) representation .
these two algorithms , minpolyquotmat and minpolyquotdef , are indeed quite simple and natural , but we want to emphasize that a careful implementation is essential for making them efficient .
the reward is performance which is dramatically better than the well known elimination approach ( see the timings in section [ sec : timings - minpoly ] ) .
there are two crucial steps for achieving an efficient implementation .
the first is when computing the powers of @xmath72 ( in step minpolyquotmat-4.1 ) and @xmath83 ( in step minpolyquotdef-4.1 ) : the _ incremental approach _ we give , using the last computed value , is very important .
the second is the search for a linear dependency ( in steps minpolyquotmat-4.2 and minpolyquotdef-4.2 ) .
we implemented it creating a c++ object called ` lindepmill ` .
this object accepts vectors one at a time , and says whether the last vector it was given is linearly dependent on the earlier vectors ; if so , then it makes available the representation of the last vector as a linear combination of the earlier vectors .
internally ` lindepmill ` simply builds up and stores a row - reduced matrix and the linear relations as new vectors are supplied .
the topic of section [ modularapproach ] is to show how to compute the minimal polynomial of an element of a zero - dimensional affine @xmath5-algebra using a modular approach . in this section
we describe the necessary tools to achieve such goal .
given an ideal @xmath0 in @xmath84 $ ] what does it mean to reduce @xmath0 modulo a prime number @xmath85 ?
since there is no homomorphism from @xmath5 to @xmath86 , there is no immediate answer to this question . in this section
we let @xmath87 $ ] , investigate the problem , and provide a useful answer .
we shall also assume that @xmath12 comes with a term ordering @xmath7 .
we start with the following lemma ( recall the notation in remark [ rem : zdelta ] ) .
[ lemma : samelocaliz ] let @xmath88 , let @xmath0 be a non - zero ideal in @xmath12 , let @xmath89 be its reduced @xmath7-grbner basis , and let @xmath18 .
assume that @xmath14 and @xmath89 have all coefficients in @xmath31 . 1 . every intermediate step of rewriting of @xmath14 via @xmath89 has all coefficients in @xmath31 .
2 . the polynomial @xmath90 has all coefficients in @xmath31 . if @xmath91 , the result is trivially true .
so we now assume @xmath92 .
if @xmath14 can be reduced by @xmath89 then there exists @xmath93 such that @xmath94 for some @xmath95 and some power - product @xmath96 .
let @xmath97 be the coefficient of @xmath98 in @xmath14 ; by hypothesis @xmath99 .
then the first step of rewriting gives @xmath100 which has all coefficients in @xmath31 .
we can now repeat the same argument for rewriting @xmath101 , and so on . using remark [ rem : nfsigma ]
, we deduce that the final result , when no further such rewriting is possible , is the normal form of @xmath14 .
it is reached after a finite number of steps .
these considerations prove both claims .
the following example illustrates the lemma .
[ ex : samelocaliz ] let @xmath102 $ ] , let @xmath103 where @xmath104 , @xmath105 , and let @xmath106 .
the reduced @xmath7-grbner basis of @xmath0 is @xmath107 where @xmath108 .
we let @xmath109 so that @xmath110 $ ] .
we have @xmath111 , and it is easy to check that the explicit coefficients in the equality @xmath112 are the coefficients of a sequence of rewriting steps from @xmath14 to @xmath90 . as shown by the lemma , they all lie in @xmath113 .
the following easy example shows that the number @xmath33 introduced in the above lemma depends on @xmath7 .
[ ex : dependondelta ] let @xmath114 $ ] , let @xmath115 where @xmath116 .
depending on the term ordering chosen , the number @xmath33 can be @xmath117 , @xmath118 or @xmath119 .
[ reductionmodp ] let @xmath33 be a positive integer , and @xmath6 be a prime number not dividing @xmath33 .
we write @xmath120 to denote both the canonical homomorphism @xmath121 and its natural `` coefficientwise '' extensions to @xmath122 \to { { \mathbb f}}_p[x_1 , \dots , x_n]$ ] ; we call them all * reduction homomorphisms modulo @xmath6*. now we need more definitions .
[ def : densigma ] let @xmath123 $ ] . 1
. given a polynomial @xmath50 , we define the * denominator of @xmath14 * , denoted by @xmath124 , to be @xmath74 if @xmath91 , and otherwise the least common multiple of the denominators of the coefficients of @xmath14 .
2 . given a non - zero ideal @xmath0 in @xmath12 , with reduced @xmath7-grbner basis @xmath89
, we define the * @xmath7-denominator of @xmath0 * , denoted by @xmath125 , to be the least common multiple of @xmath126 .
+ the following theorem illustrates the importance of the @xmath7-denominator of an ideal .
[ thm : rgbp]*(reduction modulo @xmath6 of grbner bases ) * + let @xmath0 be a non - zero ideal in @xmath127 $ ] with reduced @xmath7-grbner basis @xmath89 .
let @xmath6 be a prime number which does not divide @xmath125 . 1 .
the set @xmath128 is the reduced @xmath7-grbner basis of the ideal @xmath129 .
the set of the residue classes of the elements in @xmath130 is an @xmath43-basis of the quotient ring @xmath131/{\langle \pi_p(g ) \rangle}$ ] .
3 . for every polynomial
@xmath132 $ ] such that @xmath133 we have the equality @xmath134 .
we start by proving claim ( a ) .
every polynomial @xmath135 in @xmath89 is monic , so @xmath136 is monic and @xmath137 .
next we show that @xmath128 is a reduced @xmath7-grbner basis .
so assume @xmath138 , let @xmath139 and let @xmath140 be the @xmath141-polynomial of @xmath142 .
it rewrites to zero via a finite number of steps of rewriting : @xmath143 for @xmath144 .
let @xmath145 , then @xmath146 and every @xmath147 have all coefficients in @xmath31 .
lemma [ lemma : samelocaliz ] implies that each @xmath148 is in @xmath31 and that all coefficients of each @xmath149 are in @xmath31 .
we now show that the @xmath141-polynomial of the @xmath6-reduced pair @xmath150 rewrites to zero via the set @xmath128 .
first we see that @xmath151 . now applying @xmath120 to each rewriting step we get @xmath152 . if @xmath153 , this is a rewriting step for @xmath154 , otherwise `` nothing happens '' and we simply have @xmath155 .
this shows that all the @xmath141-polynomials of @xmath128 rewrite to zero , and hence that @xmath128 is a @xmath7-grbner basis .
finally we observe that @xmath156 for all @xmath157 , hence @xmath128 is actually the reduced @xmath7-grbner basis of the ideal @xmath129 .
as already observed , we have @xmath158 for all @xmath159 , hence claim ( b ) follows from ( a ) . for part ( c ) we let @xmath160 .
we use the same method as in the proof of part ( a ) but starting with @xmath161 .
once again all rewriting steps have coefficients in @xmath31 , and applying @xmath120 to them we get either a rewriting step for @xmath162 or possibly a `` nothing happens '' step .
therefore the image of the final remainder @xmath163 is the normal form of @xmath162 .
the following example illustrates some claims of the theorem .
[ ex : samelocalizcontinued ] we continue the discussion of example [ ex : samelocaliz ] .
we choose @xmath164 and get @xmath165 as the @xmath166-reduction of @xmath0 . from theorem [ thm : rgbp ]
we know that @xmath167 is the reduced @xmath7-grbner basis of @xmath129
. theorem [ thm : rgbp ] , in particular claim ( c ) , motivates the following definition .
if @xmath6 is a prime number which does not divide @xmath125 , then the ideal generated by the @xmath168 in the polynomial ring @xmath169 $ ] is called the * @xmath166-reduction of @xmath0 * , and will be denoted by @xmath170 .
observe that if @xmath0 is zero - dimensional so is @xmath170 .
the following example shows the necessity of considering the reduced grbner basis in theorem [ thm : rgbp ] .
[ ex : badexample ] let @xmath171 $ ] , let @xmath172 be the product of many primes , for instance the product of the first @xmath173 prime numbers , and let @xmath174 .
the set @xmath175 is a grbner basis of @xmath0 , while the set @xmath176 is the reduced grbner basis of @xmath0 .
reducing @xmath141 modulo @xmath6 where @xmath177 produces the ideal @xmath178 , while reducing @xmath89 produces the ideal @xmath179 .
more investigation about @xmath166-reductions is done in @xcite .
the topic of this section is to show how to compute the minimal polynomial of an element of a zero - dimensional affine @xmath5-algebra using a modular approach .
modular reduction is a very well - known technique , however there is no universal method for addressing the specific problems of bad reduction arising in every application .
our problem is no exception as we shall explain shortly . for more details on this topic
we recommend reading section 6 of @xcite . for this entire section the ideal @xmath0 will be zero - dimensional , and since in theorem [ thm : rgbp].(b ) we have seen that the set of power - products @xmath180 can be mapped both to a basis of @xmath84/i$ ] and also to a basis of @xmath131/i_{(p,\sigma)}$ ] , we are motivated to provide the following definition .
[ f - matrix ] let @xmath181 $ ] with term ordering @xmath7 .
let @xmath0 be a zero - dimensional ideal in @xmath12 .
let @xmath182 and let @xmath183 with elements in increasing @xmath7-order .
so the natural image of @xmath184 in @xmath1 is a @xmath5-basis of monomials for @xmath1 .
we denote the natural image of @xmath184 in @xmath84 $ ] by @xmath185 and the natural image of @xmath184 in @xmath186 $ ] by @xmath187 .
given @xmath18 we denote by @xmath188 the @xmath189 matrix whose @xmath190-th column ( for @xmath191 ) contains the coordinates of @xmath192 in the basis @xmath185 .
similarly , we denote by @xmath193 the @xmath189 matrix whose @xmath190-th column contains the coordinates of @xmath194 in the basis @xmath187 .
we observe that these matrices depend on both @xmath7 and the corresponding ideals .
the following proposition contains useful information about reduction of matrices .
[ pmultmat ] let @xmath18 be a polynomial and let @xmath195 . 1 . for every @xmath196 ,
all the entries of the matrix @xmath188 are in @xmath31 .
2 . for every @xmath196
, we have @xmath197 for any prime @xmath198 .
claim ( a ) follows from lemma [ lemma : samelocaliz ] applied to @xmath199 , and claim ( b ) follows directly from theorem [ thm : rgbp].(c ) .
we start this subsection with an elementary result which is placed here for the sake of completeness .
[ usinggauss ] let @xmath200 $ ] be monic polynomials such that @xmath135 divides @xmath14 , and let @xmath88 . if @xmath14 has coefficients in @xmath31 then also @xmath135 has coefficients in @xmath31 .
by hypothesis we have a factorization @xmath201 in @xmath202 $ ] .
set @xmath203 , @xmath204 and @xmath205 ; so each of @xmath206 , @xmath207 and @xmath208 is a primitive polynomial with integer coefficients . by gauss
s lemma @xmath209 is a primitive polynomial with integer coefficients .
hence @xmath210 ; in particular @xmath211 , and consequently @xmath212 . since @xmath213 $ ] we have @xmath214 , hence
also @xmath215 which implies that @xmath216 $ ] .
we now give a proposition which tells us which primes could appear in the denominator of a minimal polynomial .
[ goodminpoly ] let @xmath18 , and let @xmath217 .
then the minimal polynomial @xmath22 has all coefficients in @xmath31 .
let @xmath58 be the @xmath5-endomorphism of @xmath1 given by multiplication by @xmath56 .
it is known that @xmath218 ( see remark 4.1.3.(a ) of @xcite ) .
let @xmath219 be the characteristic polynomial of the endomorphism @xmath58 ; by definition @xmath220 .
next , let @xmath221 , let @xmath222 , let @xmath223 be the identity matrix of size @xmath224 , and let @xmath225 be the matrix which represents @xmath58 with respect to the basis @xmath184 .
then we have @xmath226 . the entries of @xmath225 are the coefficients of the representations of @xmath227 in the basis @xmath184 for all @xmath228 .
they are in @xmath31 by lemma [ lemma : samelocaliz ] .
so we have proved that @xmath229 $ ] . from the cayley - hamilton theorem
we deduce that @xmath230 is a divisor of @xmath219 .
it follows from lemma [ usinggauss ] that also @xmath231 $ ] .
the conclusion of the proposition above motivates the following definition .
[ def : usableprime ] let @xmath18 be a polynomial , and let @xmath6 be a prime number .
then @xmath6 is called a * usable prime for @xmath14 with respect to @xmath232 * if it does not divide @xmath233 . if @xmath0 and @xmath7 are clear from the context , we say simply a * usable prime*. a prime which is not usable is called * ugly*. it follows from the definition that , for a given input @xmath234 , there are only finitely many ugly primes , and it is easy to recognize and avoid them . in this subsection
we refine the definition of usable .
[ def : goodprime ] let @xmath6 be a usable prime for @xmath14 with respect to @xmath232 ; consequently , by proposition [ goodminpoly ] , @xmath235 is well - defined .
we say that @xmath6 is a * good prime for @xmath14 * if @xmath236 , in other words if the minimal polynomial of the @xmath6-reduction of @xmath14 modulo the @xmath166-reduction of @xmath0 equals the @xmath6-reduction of the minimal polynomial of @xmath14 modulo @xmath0 over the rationals .
otherwise , it is called * bad*. the following simple example illustrates how a prime can be bad even if it is usable .
let @xmath237 $ ] , let @xmath238 , and let @xmath239 .
the set @xmath240 is a reduced grbner basis of @xmath0 for every ordering , @xmath241 is a quotient basis of @xmath242/i$ ] regardless of ordering .
moreover we have @xmath243 regardless of ordering , and thus every prime number is usable .
over @xmath5 we have @xmath244 .
whence we deduce that @xmath245 .
if we change the base field to the finite field @xmath246 , we get @xmath247 which shows that @xmath248 .
it is easy to see that @xmath117 is the only bad prime in this case .
next we show that there are only finitely many bad primes .
[ almostallgood]*(finitely many bad primes ) * + let @xmath87 $ ] , let @xmath0 be a zero - dimensional ideal in @xmath12 , let @xmath7 be a term ordering on @xmath10 , let @xmath18 , and @xmath6 be a usable prime . 1 .
then @xmath235 is a multiple of @xmath249 .
2 . there are only finitely many bad primes .
the prime @xmath6 is good if and only if @xmath250 . to simplify the presentation we let @xmath251 and @xmath252
let @xmath253 , and set @xmath254 . by proposition [ goodminpoly ]
we have @xmath255 $ ] . by the definition of minimal polynomial we have @xmath256 .
therefore we have an equality @xmath257 for certain @xmath258 where @xmath259 is the reduced @xmath7-grbner basis of the ideal @xmath0 . by lemma [ lemma : samelocaliz ] we know that each @xmath260 .
since @xmath6 is a usable prime , it follows from proposition [ goodminpoly ] that we can apply @xmath120 to get @xmath261 which shows that @xmath262 , and hence that @xmath263 it is a multiple of @xmath264 .
so claim ( a ) is proved . to prove ( b ) and
( c ) it suffices to show that only a finite number of usable primes are such that @xmath263 is a non - trivial multiple of @xmath265 , and we argue as follows .
since @xmath196 is the degree of @xmath266 we deduce that the matrix @xmath267 has rank @xmath196 , hence there exists an @xmath268-submatrix of @xmath267 with non - zero determinant ; moreover this determinant lies in @xmath269 , so can be written as @xmath270 for some non - zero @xmath271 and some @xmath272 .
for any prime @xmath6 not dividing @xmath273 , the matrix @xmath274 has maximal rank : by proposition [ pmultmat ] we have @xmath275 .
hence for these primes the degree of @xmath264 is @xmath196 , and the conclusion follows .
we do not have an absolute means of detecting bad primes but , for two different usable primes @xmath276 and @xmath277 we can compute the minimal polynomials of @xmath278 with respect to @xmath279 , and by comparing degrees we can sometimes detect that one prime is surely bad , though without being certain that the other is good . *
( detecting some bad primes)*[cor : badprimes ] + let @xmath280 be two usable primes , and let @xmath281 and @xmath282 be the minimal polynomials of the corresponding modular reductions . 1 .
if @xmath283 then @xmath276 is a bad prime .
2 . if @xmath284 then @xmath276 is a good prime .
claim ( a ) follows from parts ( a ) and ( c ) of theorem [ almostallgood ] .
claim ( b ) follows from theorem [ almostallgood].(c ) since @xmath285 is an upper bound for the degrees of the minimal polynomials .
combining these results we get the following algorithm .
[ alg : modular ] * ( minpolyquotmodular ) * _ notation : _ : : @xmath286 $ ] with term ordering @xmath7 input : : @xmath0 , a zero - dimensional ideal in @xmath12 , and a polynomial @xmath18 1 : : compute the @xmath7-reduced grbner basis of @xmath0 2 : : choose a usable prime @xmath6 see definition [ def : usableprime ] .
3 : : compute @xmath287 and @xmath170 .
4 : : compute @xmath288 $ ] , the minimal polynomial of @xmath289 . 5 : : let @xmath290 and @xmath291 . 6
: : _ main loop : _
+ 6.1 ; ; choose a new usable prime @xmath6 .
6.2 ; ; compute the minimal polynomial @xmath292 $ ] .
6.3 ; ; if @xmath293 then + 6.3.1 : : if @xmath294 then let @xmath290 and @xmath291 .
6.3.2 : : continue with next iteration of _ main loop _ 6.4 ; ; let @xmath295 , and let @xmath296 be the polynomial whose coefficients are obtained by the chinese remainder theorem from the coefficients of @xmath297 and @xmath298 . 6.5 ; ; compute the polynomial @xmath299 $ ] whose coefficients are obtained as the fault - tolerant rational reconstructions of the coefficients of @xmath296 modulo @xmath300 . 6.6 ;
; were all coefficients `` reliably '' reconstructed ? + yes : : if @xmath301 and @xmath302 return * @xmath303 * no : : let @xmath304 and @xmath305 .
output : : @xmath299 $ ] , the minimal polynomial @xmath306 .
the correctness of this algorithm follows from theorem [ almostallgood ] and the termination from corollary [ cor : badprimes ] .
termination of the _ main loop _ depends on the test @xmath307 in step minpolyquotmodular-6.6(yes ) ; however evaluating @xmath308 modulo @xmath0 is typically computationally expensive compared to the cost of a single iteration .
for this reason , in step minpolyquotmodular-6.5 we use the fault - tolerant rational reconstruction implemented in cocoa ( see @xcite ) which gives also an indication whether the reconstructed rational is `` reliable '' ( _ i.e. _ heuristically probably correct ) .
this is a computationally cheap criterion which surely indicates `` reliable '' almost as soon as @xmath300 becomes large enough to allow correct reconstruction , while also almost certainly indicating `` not reliable '' before then .
once @xmath297 has the correct degree , the degree check in step minpolyquotmodular-6.3 ensures that only results from good primes are used ; in this situation our fault - tolerant reconstruction is equivalent to monagan s mqrr @xcite .
a disadvantage of algorithm [ alg : modular ] is that it needs a grbner basis over @xmath5 , requiring a potentially costly computation .
we can make a faster heuristic variant of the algorithm by working directly with the given generators for @xmath0 .
let @xmath309 be the set of given generators .
we shall skip all `` ugly '' primes which divide @xmath310 . in steps minpolyquotmodular-3 and
4 we use the ideal @xmath311 instead of @xmath170 . in the _ main loop _
we skip step minpolyquotmodular-6.3 , since there are no guarantees on the degrees of bad @xmath298 . in other words ,
we keep all the @xmath298 but when using the chinese remainder theorem to combine , we take only those polynomials having the same degree as the current @xmath298 . in step minpolyquotmodular-6.6(yes ) we return directly @xmath303 skipping the check that @xmath308 is in @xmath0 ( since we want to avoid computing its grbner basis )
. this heuristic algorithm may sometimes give a wrong answer : _
e.g. _ if given as input the generators in example [ ex : badexample ] in this instance the answer would be obviously wrong since it is reducible . there are only finitely many primes giving a bad @xmath298 .
we can see this by picking some term ordering @xmath7 , and tracing through the steps to compute the reduced @xmath7-grbner basis from the generators @xmath309 .
any prime which divides a denominator or a leading coefficient at any point in the computation may give a bad @xmath298 ; to these we add the ( finitely many ) bad primes for that reduced grbner basis .
all remaining primes will give a good @xmath298 .
in this section we illustrate the merits of the algorithms explained in the paper .
each example is described by introducing a polynomial ring @xmath12 , an ideal @xmath0 in @xmath12 , and a polynomial @xmath14 in @xmath12 which is denoted either by @xmath312 if it is linear or by @xmath14 if it is not linear .
the task is to compute @xmath22 , the minimal polynomial of @xmath56 in @xmath1 .
the column * example * gives the reference number to the examples listed below .
the column * gb * gives the times to compute the grbner basis ( in seconds ) ; the columns * def , mat , elim * give the times ( in seconds ) of the computation of the algorithms [ alg : minpolyquotdef ] , [ alg : minpolyquotmat ] , and [ alg : minpolyquotelim ] respectively . the column *
deg * gives the degree of the answer , as an indication of the complexity of the output . | given a zero - dimensional ideal @xmath0 in a polynomial ring , many computations start by finding univariate polynomials in @xmath0 . searching for a univariate polynomial in @xmath0
is a particular case of considering the minimal polynomial of an element in @xmath1 .
it is well known that minimal polynomials may be computed via elimination , therefore this is considered to be a `` resolved problem '' . but being the key of so many computations , it is worth investigating its meaning , its optimization , its applications .
minimal polynomial |
one of the most common problems in astrophysics is the classification of astronomical sources based on their colors and magnitudes in a given photometric system .
when high spectral resolution data ( or a large number of photometric bands covering the source sed ) are available , this classification , and the extraction of the physical parameters of the source , is generally achieved by fitting the observations with an appropriate physical model . in many instances , however , either a reliable model is not available , or the number of photometric bands is not sufficient to provide a robust source classification . when data fitting is not possible , a common fallback solution is to infer the nature of the sources to be classified by their proximity to `` regions '' in meaningful color - color and color - magnitude diagrams , where the sources of a certain class are expected to be found .
these regions are in turn defined on the basis of generic physical considerations ( e.g. stars burning h in their cores are located on the region of the herzsprung - russell diagram we call main sequence ) or by association with other sources of the same class .
a typical example of this approach in the early years of infrared space astronomy were the iras color - color diagrams @xcite aimed to automatically classify the @xmath2 sources found by the infrared astronomical satellite in its bands at 12 , 25 , 60 and 100 .
the diagrams were created by deriving the iras colors of @xmath3 sources whose nature could be inferred by the properties of their iras low resolution spectra @xcite .
the resulting diagrams were a grid of polygonal regions where sources with specific properties ( stars , circumstellar envelopes with varying degrees of optical thickness , planetary nebul and other infrared sources ) were expected to be found . as is common in these cases ,
the boundaries between the regions were defined arbitrarily by using a convenient geometrical pattern bisecting known `` template '' sources used for building the diagrams .
most importantly , these regions did not have an associated statistical meaning , e.g. it was not possible to quantify how complete and effective was the source classification provided by these regions .
other branches of science , however , have developed statistically valid techniques to attack this kind of unstructured classification problems , where detailed knowledge of a model is not required , or not available .
the k - nearest neighbors ( @xmath0-nn ) method @xcite , in particular , has been succesfully used as an efficient `` black box '' predictor for problems of pattern recognition and unsupervised machine learning , in fields ranging from computerized handwriting recognition @xcite to automatic classification of satellite imagery @xcite , to medical imaging and diagnostics . in astronomy , k - nearest neighbors methods have been traditionally used to study clustering in the spatial distribution of astronomical sources ( see e.g. @xcite ) , by analyzing the statistical distribution of the distances , on the plane of the sky or in the 3-dimensional space , between each source and its nearest neighbors . alternatively , the method has been the base of regression techniques for parameter fitting ( e.g. photometric redshifts , see @xcite ) . in this paper , we will instead apply the @xmath0-nn method in its role of nonparametric classifier , where the class of a new set of data is decided based on its _ distance _ from a class of `` templates '' , and where the distance is defined in a multi - dimensional color and magnitude space .
our implementation of the method is specifically tuned to the search of rare sources hidden in a large catalog . to illustrate the effectiveness of the method
, we apply our technique to the search of brown dwarfs with the infrared array camera ( irac , @xcite ) onboard the _
space telescope @xcite .
as shown by @xcite , brown dwarfs have unique colors in the near - ir and irac 3.6 , 4.5 , 5.8 and 8.0 bands , due to the presence of prominent molecular features such as ch@xmath4 , h@xmath5o , nh@xmath6 and co @xcite in the wavelength range covered by the camera ( see figure [ fig - spc ] ) .
these colors provide a powerful discriminant to identify brown dwarfs within the large photometric catalogs that have been produced during the _ spitzer _ cryogenic mission .
the @xmath0-nn method is particularly suited for this search , because of its high efficiency in finding `` needles in the haystack '' such as brown dwarfs , among the galactic general population and the extragalactic background .
the method is first applied using data from the _ spitzer _ extragalactic first look survey ( xfls , @xcite ) and the _ spitzer _ shallow survey @xcite , which are combined with ground based optical and near - ir surveys for further refinement of the candidate sample .
the parameter space of the possible color combinations and @xmath0-nn parameters is explored in order to provide and quantify the best possible search completeness and efficiency .
searches using only the two irac bands at 3.6 and 4.5 are also investigated , to assess the possibility of brown dwarf detection using only the two channels that will be available during the post - cryogenic _
warm mission .
section [ sec - description ] of the paper describes our implementation of the @xmath0-nn method , which is then applied in section [ sec - bd ] to search for field brown dwarfs in the xfls and shallow surveys . in section [ sec - nstars ]
the @xmath0-nn method is used to estimate the efficiency and completeness of irac photometric searches of brown dwarf companions around nearby stars . in section [ sec - concl ]
we summarize the results of these searches , and discuss other possible applications of the method .
in a typical application of the @xmath0-nn method , as described by @xcite , the class of a test element is selected by a majority vote of its @xmath0 nearest templates ( where @xmath0 is optimized to the specific problem at hand ) .
template objects for each class , ( i.e the _ training sample _ ) are distributed in the multi - dimensional space defined by the variables used in the analysis .
the class of the test element is determined by the prevalent class of the @xmath0 templates that , according to a given _ metric _ , are closer to the test element . note
that in this case the choice of the metric is only important for the selection of the @xmath0 nearest templates , after which the decision is determined by the rule adopted to weight the `` votes '' of these @xmath0 nearest templates .
the situation is however different in the classical astronomical setting , where sources of one specific class are searched among a larger set belonging to many other classes , which are generally not specified . in this case , a majority vote is not possible and a different approach is required . in order to address this challenge ,
we have developed a novel implementation of the @xmath0-nn method tuned for the search of rare sources in astronomical photometric catalogs .
our application requires templates only for the `` search class '' , relying on the assumption that the templates are an accurate representation of the class , and that the selection variables ( colors and magnitudes ) chosen in the analysis are sufficient to provide an effective discrimination .
the only quantitative criterion available for the selection , and used to determine its statistical validity , will be the final @xmath0-nn distance of each source in the input catalog , according to the chosen method .
as an example , in section [ sec - bd ] we show the application of our method to the search of brown dwarfs ( the _ signal _ ) in a catalog where most of the other sources are galaxies or regular stars in different evolutionary stages ( the _ background _ ) . in any @xmath0-nn application
the choice of the metric is arbitrary , and is ultimately determined only on the basis of its ability to effectively separate the signal from the background .
if the analysis involves @xmath7 separate variables ( in this case colors or absolute magnitudes ) , a common choice is the euclidean distance in the @xmath8-dimensional space , with the individual distances in each variable averaged ( or summed ) on the number of dimensions : @xmath9 where @xmath10 is the distance between the source @xmath11 and the template @xmath12 with respect to the variable @xmath13 .
the effect of averaging over the dimensionality @xmath7 is illustrated in figure [ fig - dist ] : the average increases the size of the ellipse enclosing test sources within a given distance from each template .
if a source has a unitary distance from a template along each variable , the distance will be larger than one in the case of a pure euclidean metric , but still equal to one in the case of averaged euclidean metric .
the latter is preferrable for multi - dimensional spaces where is not desirable to have a metric that tends to become larger as the dimensionality @xmath7 increases : in other words , it is a good choice to have a distance close to unity if the individual components of such distance in each variable are all around unity . for this reason
we have adopted the averaged euclidean distance in this implementation of the @xmath0-nn method .
furtermore , another advantage of averaging is that it allows to weight differently the individual components of the metric , in case some of the variables ( colors or magnitudes ) have a stronger discriminatory property for the problem at hand .
however , in the applications presented here we assume for simplicity that all the variables are equally important , and no weighting is necessary . to ensure that the distances along each variable play an equal role in the final metric , a normalization is required .
the metric should take into account , for example , if one color or magnitude has larger uncertainty than the others .
thus we divide each distance by its associated total uncertainty : @xmath14 where @xmath15 and @xmath16 are the statistical uncertainties of the data and templates respectively ( e.g. their 3@xmath17 photometric error ) in the variable @xmath13 , and where @xmath18 is a measure of the non - gaussian systematic errors of the template @xmath12 ( explained below ) , also in the variable @xmath13 .
the final @xmath0-nn distance of a source @xmath11 is then the ( weighted ) average of the distances to the nearest @xmath0 templates ( the optimal number of neighbors @xmath0 is determined with the techniques described in section [ ssec - apply ] ) : @xmath19 the weights @xmath20 are introduced to reduce the influence of isolated templates that happen to be much farther away than the other nearest neighbors .
a gaussian kernel is very effective for this task : @xmath21 ^ 2\ ] ] note that the gaussian kernel is parametrized with the geometric average ( on all variables @xmath13 ) of the same normalization factor @xmath22 adopted for the individual distances .
this is again necessary to scale the range of the kernel proportionally to the accuracy of the individual templates .
the extra factor @xmath0 plays the role of reducing the effectiveness of the kernel as the number of neighbors increases , which is the intended goal of using a large value for @xmath0 .
the effect of this normalization on the @xmath0-nn distance is shown in figure [ fig - norm ] .
the contours traced around the template sources enclose the areas within a given @xmath0-nn distance from the training sample .
the meaning of this region becomes obvious in the case of @xmath23 ( solid line ) .
thanks to the normalization with the total uncertainty @xmath22 , the region with a radius @xmath24 is nothing else than the union of the error ellipses around the templates . a test source within the region
will have a separation from the template class which is less than the uncertainty in the data and the templates , and will thus likely be a member of the template class . a source with @xmath25
will instead have a greater probability of not being a member , and should be rejected .
note that for @xmath26 the border of the region closely follows the location of the templates , deviating from a smooth line because of the scatter in the distribution of the training sample
. a larger value of @xmath0 , on the other end , will average on the position of the individual templates , thus producing a smoother contour , albeit at the risk of excluding isolated templates from the region .
the best value of @xmath0 will be a compromise between these two different tendencies , and needs to be determined case by case , with the parameter optimization techniques explained in the following section .
the solid line in figure [ fig - norm ] was derived without adding the non - gaussian systematic error @xmath18 in the templates . as a result , for any given @xmath0-nn distance
, the size of the region is determined only by the statistical errors in data and templates . a direct consequence of this
, however , is that templates that are separated by a distance larger than their statistical error will produce separated regions .
this is not desiderable , as in most astronomical applications the location of the templates is only partially in control of the astronomer . due to low statistics it depends on the chance location , in the color - color and color - magnitude space , of the template sources that have been possible to observe .
this is particularly true for the search of rare sources , where only a small number of templates is generally available .
the sparseness of the templates in this case may lead to a serious problem , as regions of the color - magnitude space where sources of the template class may exist could be excluded only because no templates have been observed there at the time the training sample was assembled . to correct this issue
we introduce in our @xmath0-nn metric the _ sparseness factor _
@xmath18 , which is a measure of how far apart the templates are with respect to each variable @xmath13 : @xmath27 where @xmath28 is defined as the average distance of the template @xmath12 from the remaining @xmath0 nearest templates . the dotted line ( for @xmath26 ) in figure [ fig - norm ]
shows how the introduction of the sparseness factor @xmath18 , acknowledging the inadequacy of the template distribution in a region of the parameter space , is able to reconnect the region despite the lack of templates between the two sets that are separated in the solid line region .
particular caution , however , should be taken while introducing the @xmath28 defined as in equation [ eq - sparse ] in cases where a gap in the templates distribution is actually expected in the data ( e.g. the gap in the horizontal branch for he burning giants in the herzsprung - russell diagram ) . in such cases , the presence of the gap can only be noticed , and is statistically significant , when it is sampled by a large number of templates on both sides of the gap .
if @xmath0 is chosen to be much smaller than the number of templates in the gap region , the gap will still be preserved as no template across the gap will be among the @xmath0 nearest neighbors in equation [ eq - sparse ] , and the sparseness factor will be smaller than the width of the gap . the @xmath0-nn metric we have defined in eq .
[ eq - knn ] is ideally suited for selecting rare astronomical sources based on their spectral properties from a large photometric catalog .
the first step in applying the method is the identification of the variables to use .
these variables can be any combination of colors or magnitudes . unless the number of available bands is so small that all the possible combinations can be tested ,
the best course of action is to choose the variables that , based on astrophysical considerations , provide the best discrimination ( e.g. colors sensitive to peculiar spectral features ) .
caution must be used to avoid choosing too many variables : even though one would naively assume that more variables would produce a better selection , this is not always the case and can lead to the so - called `` curse of dimensionality '' ( see e.g. @xcite ) .
also variables that do not carry a significant discriminating role should be avoided , because they would dilute the effectiveness of the metric by averaging out more effective variables .
having multiple variables sensitive to the same physical property can also be counter - productive , as it biases the metric towards this one physical characteristic , at the expense of other equally or more important discriminants .
a solution to this issue is to either to avoid adding variables not contributing any original discriminant , or to reduce their influence by fine - tuning the @xmath0-nn metric with appropriate weights .
once the variables are chosen , the @xmath29 distance of each source in the catalog can be determined .
the selection is done by comparing @xmath29 with a threshold value @xmath1 above which the sources are rejected . for maximum effectiveness the number of neighbors @xmath0 and the threshold distance @xmath1 have to be optimized for the problem at hand .
the goal of this optimization is to select the smallest possible number of candidates to follow - up , while preserving the completeness of the search . in this context
, we define the _ completeness _ @xmath30 as the fraction of the objects that are found , with respect to their expected number . in addition , we define the _ rejection efficiency _ @xmath31 as the fraction of the background objects that are successfully rejected .
an ideal search will have 100% completeness ( all sources are found ) and 100% rejection efficiency ( all the returned candidates are genuine ) . in practice both fractions will be lower than 1 , and the search parameters should be optimized to provide maximum possible rejection efficiency and completeness .
this optimization can be done using either the _ jackknife _ or the _ bootstrap _ methods @xcite . both methods attempt to estimate the statistical distribution of @xmath32 and @xmath33 to determine the best values of @xmath0 and @xmath1 .
the jackknife method tests the minimum distance for which the templates are an homogeneous and contiguous set .
given a certain @xmath0 , one measures the @xmath0-nn distance of each template from the remaining @xmath34 ( leave - one - out method ) .
once this is done for all templates , the completeness is derived as the fraction of templates that are within any given threshold distance @xmath1 . while this method is relatively straightforward , it may be very inaccurate for the search of rare sources , since the number of available templates is often small and does not cover uniformly the color and magnitude distribution of the target objects .
thus we adopt the bootstrap method . with the bootstrap method we evaluate the completeness by creating an artificial sample with the characteristics of the templates .
this artificial sample is then tested with the @xmath0-nn method to estimate how many of these simulated sources are found within a distance @xmath1 .
the test sample is generated by varying the template colors and magnitudes , adding a random offset using the statistical and systematic errors in the templates .
the statistical errors are introduced by adding a gaussian error equal to the statistical uncertainty of the templates ( @xmath16 in equation [ eq - norm ] ) .
the systematic errors are instead simulated by adding a random uniform component equal to the amplitude of the template sparseness factor @xmath18 defined in equation [ eq - sparse ] . with this method
it is possible to create an arbitrary number of simulated signal and background samples of any size , enabling the study of the statistical distribution of @xmath35 even when only a small number of templates is available . the rejection efficiency @xmath31 is similarly evaluated with the bootstrap method .
if a `` pure '' background sample can be extracted from the catalog , then it is just a matter of counting the fraction of catalog sources that are rejected with any combination of @xmath0 and @xmath1 . in most cases a pure background sample
is however not available but , if the objects in the search class are rare , any small sub - sample of the full catalog can be assumed to have a very small contamination of template - class sources .
these sub - samples can be randomized introducing gaussian noise equal to the data uncertainty @xmath15 defined in equation [ eq - norm ] .
sub - samples that by chance have a larger level of contamination will appear as outliers , and can be removed from the final distribution of @xmath36 .
once the distribution of @xmath30 and @xmath31 is known , the characteristics of the selection problem determine the value of @xmath0 and @xmath1 to choose .
if the search needs to be very selective because a large number of follow - up observations is not affordable , then the completeness can be sacrificed ( generally by using a small value of @xmath1 ) in favor of high @xmath31 . on the contrary ,
when completeness is paramount , a larger threshold distance can be adopted even though it will result in a large number of candidates . as an intermediate solution
we adopt the value of @xmath1 and @xmath0 for which @xmath37 is the highest .
after the @xmath0-nn metric is applied with the optimized parameters , all sources within the threshold distance @xmath1 should be considered as candidates .
these candidate sources can be followed - up by applying further selection criteria ( e.g. applying color cuts that could not be included in the @xmath0-nn metric , or by executing new targeted observations ) .
if the selected sample is still too big for a follow - up program , it can be helpful to run the @xmath0-nn method a second time on the first - run selection , using a different set of variables .
this is particularly effective if there is a concern that some variables have yielded a lower selection efficiency than expected , due to having been averaged out in the metric by other variables . in this case
it may be just more efficient to re - apply the @xmath0-nn method using only these variables , starting from the sources selected in the first run , rather than trying to improve the efficiency of the first @xmath0-nn run by fine tuning the weights between the variables .
as an application of the method , we present the search of brown dwarfs in _ spitzer _ surveys .
brown dwarfs represent the link between main sequence stars , fully supported by h burning in their cores , and planets . given our poor understanding of the lower end in the stellar mass distribution , an accurate census of the galactic population of brown dwarfs is of paramount importance to constrain models of star formation and galactic evolution , and to provide an accurate measurement of the stellar mass in the galaxy .
due to their low luminosity and red colors , brown dwarfs are difficult to find . the first unambiguous identification of a brown dwarf , gliese 229b @xcite , came only 20 years after the class was introduced by jill tarter in her ph.d . thesis . in recent years , however , the availability of deep wide area surveys such as the two micron all sky survey ( 2mass , @xcite ) , the deep near infrared survey of the southern sky ( denis , @xcite ) , the sloan digital sky survey ( sdss , @xcite ) and the ukirt ( uk infrared telescope ) infrared deep sky survey ( ukidss , @xcite ) allowed to identify an increasing number of brown dwarfs in the solar neighborhood ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . however , the total number of brown dwarfs known to date ( @xmath38 of the red , dusty l dwarfs , and @xmath39 of the cooler , methane rich t dwarfs , according to the dwarfarchives.org database ) is still too small to provide a reliable characterization of the substellar mass function . the sensitivity of the irac instrument onboard _ spitzer _ , and the characteristics of its photometric system , raised expectations for a large increase in the number of brown dwarfs ( especially the cooler t dwarfs ) detected using wide area _ spitzer
_ surveys .
these expectations , however , have not yet materialized .
only three t dwarfs have been discovered by _ spitzer _ : a t4.5 field dwarf in the extragalactic _ spitzer _
shallow survey @xcite , and two t dwarf companions around the nearby young stars hn peg and hd 3651 @xcite .
this state of affairs arises from the complexity of discriminating brown dwarf candidates from the large number of extragalactic red sources that are within the detection limits of irac observations .
the success of photometric searches ultimately depends on the efficiency of the selection method required to extract from these large catalogs a manageable sample of sources for spectroscopic follow - up .
this selection is usually done by applying cuts in the color and magnitude space ( see e.g. @xcite and @xcite and references therein ) .
@xcite , in particular , used a single irac color cut , @xmath40 - [ 4.5 ] \geq 0.4 $ ] , to select t dwarfs with deep ch@xmath4 absorption in the 3.6 irac band , complemented by criteria based on the photometry and the morphology of the sources in optical bands .
these criteria are not able to discriminate between brown dwarfs and high redshit quasars , and are limited to the detection of dwarfs of spectral type t3 to t6 .
the @xmath0-nn method proposed here has been designed to analyze datasets in a multi - dimensional color and magnitude space , based only on the distribution of templates without the need to define a - priori cuts .
thus it is in principle capable to go beyond these limitations , opening the search to l and early - t dwarfs and t dwarfs of type later than t6 , without introducing the biases associated with the choice of the cuts , and provide a more efficient and complete search .
this will be especially important during the _ spitzer _ warm mission planned from the spring of 2009 when , upon exhaustion of the cryogenic lhe , the observatory will be tasked to conduct large area surveys using only the irac bands at 3.6 and 4.5 @xcite .
in this section we study how to improve on the selection efficiency of _ spitzer_/irac search of brown dwarfs using the @xmath0-nn method .
we explore the @xmath0-nn parameter space to understand the best strategy for these searches , and what are the requirements , in terms of auxiliary data , for their success .
we also develop techniques that allow to assess the completeness of the result , necessary to draw statistically valid conclusions from these searches . as test datasets to illustrate our @xmath0-nn search for brown dwarfs , we use two publicly available _
spitzer_/irac wide field surveys : the xfls @xcite and the botes field of the irac shallow survey @xcite .
the xfls main field covers an area of 3.8 deg@xmath41 at high galactic latitude , observed to a sensitivity reaching a @xmath42 vega magnitude of 18.9 , 18.0 , 15.7 and 15.1 at 3.6 , 4.5 , 5.8 and 8.0 respectively ( obtained with integration times of at least 60 sec per pointing ) .
the xfls main field was chosen for the availability of extensive auxiliary data , including sdss and 2mass .
the botes field of the shallow survey instead covers an area of 8 deg@xmath41 with limiting @xmath42 vega magnitudes of 18.4 , 17.7 , 15.5 and 14.8 at 3.6 , 4.5 , 5.8 and 8.0 respectively ( integration time @xmath43 sec ) .
deep near - ir @xmath44 an @xmath45 data has been obtained as part of the flamingo extragalactic survey ( flamex , @xcite ) for 7.1 deg@xmath41 of the irac field , and in the optical as part of the noao deep wide - field survey ( ndwfs , @xcite ) .
the main difference between the two samples ( apart for the shallow survey being almost twice the area of the xfls ) , is in the depth of the optical and near - ir ancillary data . while sdss provides @xmath42 detection limits of 22.3 , 23.3 , 23.1
, 22.3 and 20.8 in @xmath46 , @xmath47 , @xmath48 , @xmath49 and @xmath50 @xcite , ndwfs has @xmath42 point source depths of 27.1 , 26.1 and 25.4 in @xmath51 , @xmath52 and @xmath53 respectively @xcite .
2mass provides @xmath42 sensitivity in @xmath44 , @xmath54 and @xmath45 of 16.6 , 15.9 and 15.1 respectively @xcite , while flamex approaches a @xmath42 sensitivity of 21.4 and 19.5 in @xmath44 and @xmath45 @xcite .
the added depth of the ndwfs and flamex data provides a powerful tool to resolve ambiguities between red sources in the irac bands of galactic and extragalactic nature . by testing our method on both datasets we can measure the efficiency of the brown dwarf search on surveys with different depth and assess the auxiliary data requirements necessary to enable effective brown dwarf searches during the _ spitzer _ warm mission .
figure [ fig - fls ] shows the distribution of point sources from the xfls in a number of irac and 2mass colors , compared with the distribution of 37 l , 7 early - t ( [email protected] ) and 22 late - t ( [email protected] ) templates from @xcite .
we chose the colors in the diagrams to maximize the separation between brown dwarfs and other galactic ( clump near zero colors ) and extragalactic ( long plume with red colors ) sources .
in particular ( see figure [ fig - spc ] ) : ( 1 ) the @xmath57 $ ] color is effective in separating the l dwarfs from regular stars , mainly due to h@xmath5o absorption in the @xmath44 band ; ( 2 ) the @xmath58 $ ] and @xmath40-[4.5]$ ] colors separate the t dwarfs from all other sources ( l dwarfs , regular stars and extragalactic objects ) , due to the increasing ch@xmath4 absorption in the @xmath59 and 3.6 bands ; ( 3 ) the @xmath60-[5.8]$ ] color is useful to select again the t dwarfs , which appear blue because of h@xmath5o absorption in the 5.8 band ; and ( 4 ) the @xmath40-[8.0]$ ] color is also providing a strong separation of the t dwarfs due to the methane absorption which is stronger at shorter wavelength than in the reddest irac band .
color combinations using the @xmath54 band have a similar discriminatory power than colors with the @xmath44 and @xmath45 photometry : @xmath61 $ ] shows a trend analogous to the @xmath57 $ ] color , and @xmath62 $ ] has a very similar color trend than @xmath58 $ ] .
the @xmath63 colors provide a similar discrimination than the @xmath40-[4.5]$ ] colors , because of ch@xmath4 absorption stronger in the @xmath45 than the @xmath44 band .
the late - t dwarfs appear well separated from any other source , thanks especially to the @xmath40-[4.5]$ ] irac color , even though some contamination persists with red high - redshift quasars having pah emission .
the early - t and l dwarfs are however more difficult to discriminate because , once dispersion due to photometric errors is taken into account , their color space is very similar to the one occupied by low - redshift galaxies and regular stars .
for this reason , and due to the relative small number of early - t dwarfs in the @xcite sample , our @xmath0-nn selection is done for two separate classes only : one comprising all l and early - t templates , and one with late - t ( with [email protected] ) dwarfs . to explore the effect of the presence or absence of individual bands in the selection efficiency , we have divided our xfls and shallow survey catalogs in 2 different subsamples : ( 1 ) sources having @xmath64 photometry in @xmath44 , @xmath45 and all four irac bands ; ( 2 ) sources having @xmath64 detection in @xmath44 , @xmath45 , 3.6 and 4.5 .
the first sample is intended to test the effectiveness of the @xmath0-nn method when all ir bands are available ( the @xmath54 band has not been considered because of its unavailability in flamex , and because brown dwarf colors using the @xmath54 band are very similar to the colors using @xmath44 and @xmath45 ) . the second sample is instead designed to simulate the case of the _ spitzer _ warm mission , when only the two short wavelength irac channels will be available , and also to avoid the limitations imposed by the less sensitive 5.8 and 8.0 bands in currently available surveys . for each sample
the search is done using a subset of the available colors , avoiding the repetition of similar colors , that would dilute the @xmath0-nn metric .
the characteristic of the individual subsamples , their size and the color combinations used in the @xmath0-nn search are listed in table [ tab - samples ] .
optical photometry and imaging are not used at this stage , because only a limited number of the brown dwarfs we are using as templates have reliable magnitudes at optical wavelengths .
the optical data will however be crucial to refine the search results later on .
the best values of @xmath0 and @xmath1 for the search can be determined by using the bootstrap method described in section [ ssec - apply ] .
the goal is to optimize the @xmath0-nn parameters in order to have the smallest possible number of candidates that will need follow - up observations while preserving the completeness of the search .
the jackknife method is not suitable in this case because of the very small number of available templates in each brown dwarf class .
the completeness of the search in this case is @xmath65 , where @xmath66 is the number of brown dwarfs expected to exist in the dataset and @xmath67 is the number of false negatives ( i.e. true brown dwarfs not identified ) in the search .
the rejection efficiency can be written as @xmath68 where @xmath69 is the fraction of false positives , i.e. the number @xmath70 of incorrectly identified brown dwarfs , with respect to the total number of sources @xmath71 in the sample . figure [ fig - opt ] shows the rejection efficiency and completeness for a monte carlo test using 100 randomized samples each with 500 background sources and 500 simulated brown dwarfs , for @xmath72 , 5 and 7 .
the simulations are made for the subsample using irac plus @xmath44 and @xmath45 bands ( for both l / early - t and late - t searches ) , and for _ spitzer _ warm mission colors ( as described in table [ tab - samples ] ) .
the figure shows the resulting rejection efficiency and completeness for the xfls ; similar results are obtained for the shallow survey . the rejection efficiency curve is a decreasing function of @xmath1 because when sources with larger @xmath0-nn distances are selected it is more likely to include false positives in the candidates . on the other hand , for smaller @xmath1
more true brown dwarfs are missed , leading to a smaller completeness .
we have adopted the @xmath1 for which the two curves cross .
the values of @xmath1 at the @xmath31 and @xmath30 crossing point for @xmath73 , 5 and 7 are listed in table [ tab - ec ] for simulations of the xfls and shallow survey search subsamples .
note that @xmath1 tends to be smaller for l / early - t than for late - t searches .
this is because ( as shown in figure [ fig - opt ] ) the efficiency @xmath31 of l / early - t searches drops faster with @xmath1 than in late - t searches , due to the irac colors of l / early - t dwarfs being relatively similar to the colors of regular late spectral type stars and low redshift galaxies ( see figure [ fig - fls ] ) : in l / early - t searches even a small increase in @xmath1 results in a large contamination of background sources and thus in a fast drop of the efficiency @xmath31 . as a result , in l / early - t searches the crossing point @xmath74 occurs for smaller values of @xmath1 than in late - t searches where , thanks to the unique colors of late - t dwarfs , the efficiency drops slowly as a function of @xmath1 .
table [ tab - ec ] shows that the search of late - t dwarfs using irac and near - ir colors combined reaches high level of completeness and rejection efficiency , up to 99.9% . even if only the two short wavelength irac bands are used ( as will be the case in the warm mission ) , @xmath31 and @xmath30 are still similarly high .
note however that the completeness and rejection efficiency of the warm mission search tend to be smaller when irac photometry is combined with the deeper flamex dataset , rather then the shallower 2mass , due to the higher number of red extragalactic sources cross - correlated with the irac catalog .
the l / early - t search is less efficient than the late - t search , with @xmath30 and @xmath31 @xmath75% , due to the higher contamination of this sample with background sources with similar colors .
rejection efficiency and completeness are generally higher for small @xmath0 , because in that case the selection region follows more closely the distribution of the templates . using small @xmath0
, however , puts us at risk of depending critically on outliers in the template class .
we adopted @xmath76 for the searches done using all irac bands . for the search in `` warm mission '' condition , however , we adopted @xmath72 to have the maximum possible efficiency , given the larger size of the initial catalog .
the results of the @xmath0-nn search for brown dwarfs are presented in table [ tab - knn - sel ] .
the table shows , for each subsample of the xfls and shallow survey , the number @xmath77 of selected sources .
the actual rejection efficiency @xmath78 ( where @xmath79 is the number of sources in each subsample ) is also shown . comparison with table [ tab - ec ] shows that @xmath80 is very similar to the rejection efficiency @xmath81 predicted by the bootstrap method monte carlo simulations .
table [ tab - knn - sel ] shows that , when all irac bands are available , the @xmath0-nn method alone is efficient enough to reduce the number of possible candidates to a size small enough to allow visual inspection of the candidates . as only the @xmath0-nn method
is used , the selected candidates are not biased by the choice of arbitrary cuts , depending only on a metric with known completeness , normalized on the statistical uncertainties of both data and templates .
optical data are however available , and can be used to further reduce the candidate sample , which is still necessary in the case of l / early - t dwarf searches , and when requiring only 3.6 and 4.5 @xmath82 m photometry ( resulting in a much larger catalog ) .
the low temperature of brown dwarfs requires their optical colors to be very red , in contrast with the case of extragalactic objects that tends to have a flatter optical sed .
in particular , according to @xcite , l dwarfs have @xmath83 in the sdss photometric system , while t dwarfs have @xmath84 . according to @xcite l and early - t dwarfs
have @xmath85 while according to @xcite late - t dwarfs must have @xmath86 in the ndwfs bands .
these criteria can be applied to all selected sources , as long as their @xmath49 and @xmath50 magnitudes ( for sources in the sdss ) or @xmath52 and @xmath53 magnitudes ( for ndwfs sources ) are known .
sources missing optical photometry can still be considered as brown dwarf candidates , given that their absence from the optical catalog may be an indication of very red colors .
these colors can not be introduced directly in the @xmath0-nn metric because good photometry in the optical bands is missing for a significant fraction of our templates . by using the optical bands in the form of `` cuts '' we are introducing a bias associated to the choice of the cut .
however this is still more efficient that doing the selection using cuts alone , since the candidates have been pre - selected in an unbiased mode using all the other bands with the @xmath0-nn metric .
this allows us to adopt a less stringent optical color criteria while still preserving very high efficiency in the selection .
it also allows to retain the candidates selected with the @xmath0-nn method that are missing optical data ( the potentially coolest late - t candidates ) , which would be eliminated by default if color cuts were the only selection criteria .
the application of the criteria described above reduces drastically the number of viable candidates .
the remaining candidates can then be checked visually from the sdss and ndwfs plates , to eliminate all sources that appear extended , that are blended with other sources , or that are corrupted by artifacts in the images .
the remaining candidates are listed in table [ tab - xfls ] and [ tab - flamex ] .
the number of viable candidates after the optical criteria are applied is listed in table [ tab - knn - sel ] , separately for the @xmath87 sources that possess optical detection , and the @xmath88 that are not detected in the optical surveys .
note that when all irac , @xmath44 and @xmath45 photometry is known , cross - checking with optical data reduces the number of late - t dwarf candidates to zero for the xfls , and to only one for the shallow survey .
the lone t dwarf selected for the shallow survey is in fact the t4.5 dwarf irac j142950.8@xmath89333011 found by @xcite in their search .
this result shows that the @xmath0-nn method , when used on near - ir and _ spitzer _ data , in combination with deep optical photometry is capable to select true late - t dwarfs while rejecting all other sources of different nature .
in fact , the method here described effectively rejected the second red source found by @xcite color cuts , the @xmath90 radio - loud quasar irac j142738.5@xmath89331242 .
figure [ fig - sel-7b ] shows the @xmath91 $ ] and @xmath40 - [ 4.5]$ ] colors of all the candidates selected in both surveys . of the plotted sources , 5 have colors at odd with the brown dwarf templates .
these sources , most likely red background galaxies , have identical colors than t dwarfs in all bands , with the exception of the two plotted here .
the fact that they have been selected by the @xmath0-nn method is an example of `` variable dilution '' in the metric , as described in section [ ssec - apply ] . to refine the selection
we can apply a second time the @xmath0-nn method , using only this pair of variables , the @xmath92 $ ] and @xmath40 -[4.5]$ ] colors .
the region corresponding to @xmath93 and @xmath94 ( providing completeness @xmath95% in these two variables ) is plotted in figure [ fig - sel-7b ] , and confirms that these sources are outliers .
we flagged them as such in table [ tab - flamex ] .
after these anomalies are taken into account , we are left with 1 viable l / early - t and no viable late - t candidates in the xfls ( table [ tab - xfls ] ) , and 3 viable l / early - t candidates and 1 late - t dwarf ( the t4.5 dwarf j142950.5@xmath89333011 ) in the shallow survey ( table [ tab - flamex ] ) .
the l / early - t viable candidates need follow - up observations ( currently in progress ) to determine unambiguously their nature .
as a second example illustrating the case of @xmath0-nn metric using not only colors but also absolute magnitudes , we show the case of the search of low mass companions around nearby stars .
this search was conducted as part of the irac guaranteed time observations ( p.i .
giovanni fazio ) programs pid 33 , 34 , 35 , 36 , and 48 @xcite .
the survey focused on 400 stars within 30 pc from the sun , among which all stars and brown dwarfs within 5 pc known to date .
the sample included young stars with age @xmath96 myr , stars with known radial velocity discovered exoplanets , and the l and t dwarfs from @xcite used here as templates .
each star was imaged to a depth of of @xmath97 sec in all irac bands , with a field of view of @xmath98 arcmin , sufficient to detect companions at a distance of @xmath99 to 4,000 au from the primary , to a limiting mass of @xmath10020 m@xmath101 .
the search has yielded the discovery of two new t dwarfs from the whole sample , presented in @xcite .
all the observations were made in a single epoch , preventing the search of companions by virtue of their common proper motion with the primary ( except for the few cases in which a deep near - ir image was available ) .
the candidate selection was instead based on the @xmath0-nn method described below . while a paper analyzing the search results is in preparation ( carson et al . , 2009 ) , we want here to discuss the efficiency of the @xmath0-nn method in this particular case of brown dwarf search .
figure [ fig - mag - ns ] shows simulated @xmath60 $ ] vs. @xmath40-[4.5]$ ] photometry of l , early - t and late - t dwarfs around a nearby star at 5 , 10 and 20 pc . to simulate the photometry of background sources around the primary star
, we have used the xfls catalog , which is an adequate representation of a background field projected at high galactic latitude . for nearby stars projected closer to the plane of the galaxy
, the proportion of extragalactic / galactic sources will be smaller , reducing the contamination from red extragalactic sources , while contamination from red galactic sources is likely to be increased .
the latter ( mass losing evolved stars and young stellar objects ) may however be discriminated by means of auxiliary infrared observations capable of detecting the thermal emission from their circumstellar dust .
figure [ fig - mag - ns ] also shows @xmath0-nn regions plotted for @xmath76 and @xmath102 . by adding the @xmath60 $ ] magnitude to the @xmath0-nn
variables the selection is in principle improved because a large fraction of the red background extragalactic sources are fainter than the expected brightness of brown dwarf companions .
this is especially true for t dwarfs around stars within 510 pc from the sun .
this advantage is reduced for searches around farther stars , since the brightness of t dwarf companions at @xmath103 pc is the same of the extragalactic background .
we have estimated the rejection efficiency and completeness of this search using the same bootstrap method described in section [ ssec - apply ] .
table [ tab - ec - ns ] shows that for early - t dwarfs the selection efficiency reaches very high values ( up to 99% for primaries at 5 pc ) . for
late - t dwarfs the rejection efficiency is approximately the same that is obtained by using only colors as selection criteria , and the inclusion of absolute magnitudes does not result in a dramatic improvement in the search effectiveness ( suggesting that the @xmath0-nn color selection is already close to maximum efficiency ) .
the @xmath31 and @xmath30 obtained in the simulations , are adequate for this kind of search : the typical number of sources in the @xmath104 stars part of the irac companion search program had an average of @xmath105 background sources each . with this efficiency , for each field
the chosen @xmath0-nn criteria selected up to 3 candidates , many of them actually present in at least one 2mass map , and could be ruled out either by the absence of proper motion , or because they did not possess the correct near - ir colors .
the few remaining candidates were followed - up spectroscopically and by acquiring deep near - ir images , resulting in the two new t dwarfs found around hn peg and hd 3651 , presented in @xcite .
based on the photometry given by @xcite , the sensitivity of the xfls and the shallow survey in the irac bands allows for the detection of late - t dwarfs ( t4 to t8 spectral type ) up to a distance of @xmath106 pc .
this limit is determined mainly by the lower sensitivity of the 5.8 and 8.0 bands , for the latest spectral types .
if only the 3.6 and 4.5 bands are used , as in the _ spitzer _ warm mission , the higher sensitivity allows detection of late - t dwarfs up to a distance of @xmath99 pc .
the detection limit at the l - t boundary is @xmath107 pc ( @xmath108 pc if only the 3.6 and 4.5 bands are required ) . within this volume ( corrected for the malmquist bias )
, we can expect a brown dwarf search to be as complete as @xmath109% , as estimated in section [ ssec - optimization ] ( multiplied by the completeness of the original survey , and corrected for binarity ) .
searches using 2mass photometry , however , will have more stringent limits , of @xmath110 pc for late - t dwarfs and @xmath111 pc at the l - t boundary .
the sensitivity of the flamex survey is such that any l or t dwarf detected in the irac 3.6 and 4.5 bands will also be detected in the @xmath44 band , even though the lower sensitivity of the @xmath59 band limits the maximum detection distance for a t8 dwarf to @xmath112 pc .
the very small number of brown dwarf candidates that are not optically detected in at least the @xmath53 band shows that the depth of the ndwfs is not a significant constraint in the search of brown dwarfs .
the main limitation rather comes from the depth of the irac data .
these considerations come into play to understand the potential for brown dwarf searches in the recently approved exploration science surveys in the _ spitzer _ warm mission .
the requirements for brown dwarf detection are clearly a large survey area , deep observations and the availability of matching near - ir and possibly optical data . of the approved warm mission programs , three
satisfy these requirements : the `` _ spitzer _ extragalactic representative volume survey ( servs ) '' ( pi mark lacy , pid 60024 ) , the `` _ spitzer _ extended deep survey ( seds ) '' ( pi giovanni fazio , pid 60022 ) and the `` glimpse360 : completing the _ spitzer _ galactic plane survey '' ( pi barbara whitney , pid 60020 ) .
our analysis shows that when near - ir data of sufficient depth are available ( as in the case of the flamex survey ) , the search for late - t dwarfs using the photometric @xmath0-nn technique described in section [ sec - bd ] is extremely efficient and complete ( more than 99.8% @xmath31 and @xmath30 ) .
once a single @xmath113 or @xmath114 optical color is applied , the final selection produces a very small number of viable candidates to be checked individually ( 4 l / early - t and 1 late - t viable candidates ) .
it is worth noting that the only late - t dwarf candidate selected in our search is indeed a t4.5 dwarf , as discovered by @xcite . with @xmath115% completeness
we can assert that this is the only late - t dwarf present in the volume of the survey ( 7.1 deg@xmath41 for the flamex field with a depth of @xmath116 pc ) , even when only the two short wavelength irac bands are considered .
this number is consistent with the results from the t dwarf ukidss dr2 large area survey @xcite , that estimates @xmath117 late - t dwarfs in an area of 280 deg@xmath41 for a depth of @xmath118 ( corresponding to a maximum distance of @xmath119 pc for t8 dwarfs ) . by factorizing the search volume between the two surveys , where @xmath120
are the survey areas and @xmath121 their search distance limit ] one predicts @xmath1223 late - t dwarfs in the volume of the shallow survey analyzed in this paper .
this is consistent with our result ( 1 viable candidate in the flamex search ) , upon considering poisson statistics . the number of l / early - t candidates that we found shows that , with @xmath123% completeness , @xmath124 l / early - t dwarfs are present in the shallow survey searchable volume ( 8 deg@xmath41 @xmath125 60 pc ) .
these numbers can be used to estimate the potential yield of brown dwarfs in the approved warm _
exploration science surveys
. the most promising _
warm mission project for the application of the @xmath0-nn method here described is the servs survey . with a total area of 18 deg@xmath41 , and a total exposure time of 600 sec per pointing
, it will probe a brown dwarf volume almost 9 times larger than the irac shallow survey ( search depth of @xmath126 pc for late - t dwarfs scales as the limiting flux @xmath127 , in turn scaling as the signal - to - noise ratio s / n @xmath128 , where @xmath129 is the exposure time .
this gives @xmath130 and thus @xmath131 ) .
a large fraction of the survey will overlap with the video vista survey , providing a depth of 25.7 , 24.6 , 24.5 , 24.0 and 23.5 mags in the @xmath132 , @xmath133 , @xmath44 , @xmath54 and @xmath59 bands , more than matching the depth of servs in the irac 3.6 and 4.5 bands . based on these sensitivity
, servs may find as many as @xmath134 late - t dwarfs and a large number of l and early - t dwarfs .
the seds program , on the other hand , has a much smaller survey area ( @xmath135 deg@xmath41 ) but a much longer integration time ( 12 hours per pointing ) .
this scales down to a searchable volume of @xmath136 times the total volume of the shallow survey ( search depth of @xmath137 pc for late - t dwarfs ) .
this survey can potentially provide as many as @xmath138 late - t dwarfs , even though a decrease in the brown dwarf density should be expected as the survey probes farther distances from the galactic mid - plane .
this survey , however , may be limited by the challenge of finding ancillary optical and near - ir data matching the depth of the irac photometry .
while this may limit the effective late - t dwarf searchable volume , the potential search depth offered by the irac data , in a high galactic latitude region , will provide an important test for the vertical distribution of the brown dwarf population in the galaxy .
the glimpse360 survey , instead , compensates the rather shallow coverage ( 36 sec integration time for each pointing ) with a very large survey area ( 187 deg@xmath41 ) . a substantial portion of this area
is also covered by the ukidss survey @xcite , implying that @xmath139 of the 17 late - t dwarfs estimated for the whole ukidss may be present in the glimpse360 area . according to the glimpse360 consortium , more detailed simulations by @xcite predict 70 t0 , @xmath105 t5 and @xmath14020 t8 dwarfs in the survey search area . the challenge will be to isolate these brown dwarfs from the high - confusion galactic field , and distinguish them from other red galactic sources
. the @xmath0-nn method can play an important role for this task .
the main advantage of the @xmath0-nn method presented in this paper is that it allows to perform photometric searches using a large number of color and magnitude variables , defining complex regions that closely follow the distribution of the sources to be selected .
while similar regions can be defined manually , our method prevents the introduction of biases in the selection due to the choice of the cuts . with the @xmath0-nn method the selection regions in the color / magnitude space
are only based on the photometric properties of the search class and the statistical uncertainties of the data .
also , the @xmath0-nn search can be controlled by just two parameters ( the number of neighbors @xmath0 and the threshold distance @xmath1 ) , instead of many arbitrary cuts , which allows to quickly experiment different combination of variables , and optimize the search for maximum rejection efficiency and completeness .
the examples presented in this paper show that the @xmath0-nn method is an effective procedure for the search of field and companion brown dwarfs in
_ spitzer _ wide field surveys .
this can be an important asset for the _ spitzer _ warm mission surveys .
as these surveys will image areas of the sky where deep photometric catalogs in the optical and near - ir are already available , or in progress , the @xmath0-nn method can effectively select t dwarf candidates , potentially leading to a significant increase in the known number of members in this class .
this is by no means the only potential application for this method .
the method is general enough to allow the photometric selection of sources of any kind , as long as a sample of templates is available .
if enough classes of templates are used , the @xmath0-nn method can be the engine for the photometric classification of all sources in a survey , by attributing to each source the class with the highest @xmath0-nn score .
we are currently applying this method of classification to the point source catalog of the sage survey @xcite .
this work is based in part on observations made with the _ spitzer _ space telescope , which is operated by the jet propulsion laboratory , california institute of technology under a contract with nasa .
this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation .
it also used data from the sloan digital sky survey ( see full acknowledgment at http://www.sdss.org/collaboration/credits.html ) , and software provided by the us national virtual observatory , which is sponsored by the national science foundation .
we thank the _ spitzer _ shallow survey and flamingo extragalactic survey ( flamex ) teams for permission to use data from their surveys .
we also thank the national optical astronomy observatory ( noao ) deep wide - field survey team for providing the optical and near - ir imaging data used in the botes field .
noao is operated by the association of universities for research in astronomy ( aura ) , inc .
, under a cooperative agreement with the national science foundation .
the authors would finally like to thank peter eisenhardt , daniel stern , mark brodwin and buell jannuzi for useful discussions and suggestions , and the anonymous referee for insightful comments that helped improving this manuscript .
bahcall , j. n. , soneira , r. m. 1981 , , 246 , 122 ball , n. m. , brunner , r. j. , myers , a. d. , strand , n. e. , stacey , l. alberts , tcheng , d. & llor , x. 2007 , 663 , 774 burgasser , a. j. et al .
2002 , , 564 , 421 burgasser , a. j. , kirkpatrick , j. d. , mcewain , m. w. , cutri , r. c. , burgasser , a. j. , skrutskie , m. f. 2003 , , 125 , 850 burgasser , a. j. 2004 , , 155 , 191 burgasser , a. j. et al .
2006 , , 166 , 585 burrows , a. , sudarsky , d. , hubeny , i. 2006 , , 640 , 1140 chiu , k. et al .
2006 , , 131 , 2722 cruz , k. l. , reid , i. n. , liebert , j. , kirkpatrick , j. d. , lowrance , p. j. 2003 , , 126 , 2421 cruz , k. l. et al .
2007 , , 133 , 439 cushing , m. c. , rayner , j. t. & vacca , w. d. 2005 , , 623 , 1115 dahn , c. c. 2002 , , 124 , 1170 eisenhardt , p. r. et al .
2004 , , 639 , 816 elston , r. et al .
2006 , , 639 , 816 epchtein , n. et al .
1999 , , 349 , 236 fazio , g. g. et al .
2004 , , 154 , 10 fix , e. , hodges , j. 1951 , technical report 21 - 49 - 004 , 4 us air force , school of aviation medicine , randolph field , tx geballe , t. r. et al .
2002 , , 564 , 466 jannuzi , b. t. & dey , a. 1999 , in asp conf .
191 , photometric redshifts and high - redshift galaxies , ed .
r. weymann et al .
( san francisco : asp ) , 111 hastie , t. , tibshirani , r. & friedman , j. 2003 , `` the elements of statistical learning '' , 2nd edition , springer verlag , isbn : 978 - 0 - 387 - 95284 - 0 hechenbichler , k. , schliep , k. p. 2004 , sfb 386 , discussion paper 399 , ludwig - maximilians university , munich kirkpatrick , j. d. et al . 2000 , , 120 , 447 kirkpatrick , j. d. 2005 , , 43 , 195 knapp , g.r .
2004 , , 127 , 3553 macy , m. et al .
2005 , , 161 , 41 lawrence , a. et al , 2007 , , 370 , 1599 leggett , s. k. et al .
2000 , , 536 , l35 leggett , s. k. et al .
2002 , , 564 , 452 iras catalogues and atlases , atlas of low resolution spectra , iras science team 1986 , , 65 , 607 looper , d. l. , kirkpatrick , j. d. & burgasser , a. j. 2007 , , 669 , 97 luhman , k. l. et al .
2007 , , 654 , 570 meixner , m. et al . 2006 , , 132 , 2268 michie , d. spiegelhalter , d. , taylor , c. 1994 , in `` machine learning , neural and statistical classification '' , ellis horwood series in artificial intelligence , ellis horwood oppenheimer , b. r. , kulkarni , s. r. , matthews , k. & van kerwijk , m. h. 1998 , , 502 , 932 oppenheimer , b. r. , kulkarni , s. r. , matthews , k , van kerkwijk , m. h. 1995 , science 270 , 1478 patten , b. m. et al .
2005 , in `` protostars and planet v '' , proc . of the conference held october 24 - 28 , 2005 , hilton waikoloa village , the big island , hawaii , lpi contribution 1286 , p. 8042
patten , b. m. et al .
2006 , , 651 , 502 pinfield , d. j. et al .
2008 , , in press , arxiv:0806.0294 roelling , t. l. et al .
2004 , , 154 , 418 simard , p. , le cun , y. , denker , j. 1993 , in `` advances in neural information processing systems '' , morgan kaufman , san mateo , ca , p. 50
skrutskie , m. f. et al .
2006 , , 131 , 1163 stern , d. et al .
2007 , , 663 , 677 storrie - lombardi , l. j. & silbermann , n. a. 2007 , `` the science opportunities for the warm spitzer mission workshop '' , american institute of physics conference proceedings , vol .
943 tarter , j. c. 1975 , ph.d .
thesis ( university of california , berkeley ) tinney , c. g. , burgasser , a. j. , kirkpatrick , j. d. & mcelwain , m. w. 2005 , , 130 , 2326 van der veen , w. e. c. j. , habing , h. j. 1988 , , 194 , 125 werner , m. w. et al .
2004 , , 154 , 1 york , d. g. et al .
2000 , , 120 , 1579 lrl xfls irac/2mass & 4,552 & @xmath40-[4.5]$ ] , @xmath40-[8.0]$ ] , @xmath60-[5.8]$ ] , @xmath57 $ ] , @xmath58 $ ] + xfls irac warm & 8,133 & @xmath40-[4.5]$ ] , @xmath57 $ ] , @xmath58 $ ] , @xmath63 + shallow survey irac / flamex & 15,847 & @xmath40-[4.5]$ ] , @xmath40-[8.0]$ ] , @xmath60-[5.8]$ ] , @xmath57 $ ] , @xmath58 $ ] + shallow survey irac warm & 71,590 & @xmath40-[4.5]$ ] , @xmath57 $ ] , @xmath58 $ ] , @xmath63 + lccccccccc @xmath72 & & 0.62 & 93.1% & & 0.72 & 99.7% & & 0.68 & 99.3% + @xmath76 & & 0.74 & 90.3% & & 0.89 & 98.9% & & 0.87 & 98.3% + @xmath141 & & 0.83 & 87.7% & & 1.05 & 97.5% & & 1.06 & 95.9% + + @xmath72 & & 0.56 & 89.7% & & 0.76 & 99.9% & & 0.63 & 97.8% + @xmath76 & & 0.68 & 85.2% & & 0.97 & 99.8% & & 0.81 & 95.3% + @xmath141 & & 0.76 & 83.3% & & 1.17 & 99.4% & & 0.99 & 92.2% + lrrrrrrr l / early - t & 5 & 0.74 & 4,552 & 455 & 90.0% & 1 & 2 + late - t & 5 & 0.89 & 4,552 & 45 & 99.0% & 0 & 0 + late - t ( warm ) & 3 & 0.68 & 8,133 & 17 & 99.8% & 0 & 0 + + l / early - t & 5 & 0.68 & 15,847 & 2,831 & 82.1% & 3 & 1 +
late - t & 5 & 0.97 & 15,847 & 40 & 99.7% & 1 & 0 + late - t ( warm ) & 3 & 0.63 & 71,590 & 1,582 & 97.8% & 4 & 1 + rllrrrrrrrll 1 & 260.38831 & @xmath8959.27060 & 16.36 & 15.32 & 14.70 & 14.33 & 14.36 & 14.18 & 13.93 & l / early - t & red star ?
+ 2 & 261.10120 & @xmath8960.03591 & 15.74 & 15.06 & 14.77 & 13.96 & 13.97 & 14.03 & 14.00 & l / early - t & + 3 & 261.13129 & @xmath8960.05562 & 16.59 & 16.05 & 15.20 & 14.59 & 14.42 & 14.52 & 14.36 & l / early - t & galaxy ?
+ lllrrrrrrrrrll 1 & 216.278002 & @xmath8934.355660 & & & 23.48 & 19.53 & 19.07 & 17.63 & 17.34 & & & late - t & red galaxy ? + 2 & 216.603543 & @xmath8934.140991 & & & 23.51 & 20.59 & 19.41 & 18.57 & 17.74 & & & late - t & red galaxy ?
+ 3 & 217.032547 & @xmath8934.098458 & & & & 20.32 & 19.16 & 18.07 & 17.58 & 16.64 & & late - t & red galaxy ?
+ 4 & 217.462015 & @xmath8933.503213 & & & 22.21 & 16.88 & 16.99 & 15.70 & 15.12 & 15.21 & 14.59 & t4.5 & j142950.9@xmath89333011 + 5 & 217.786508 & @xmath8933.139283 & & 22.75 & 20.40 & 17.38 & 16.30 & 15.80 & 15.69 & 15.35 & 15.12 & l / early - t & + 6 & 217.920577 & @xmath8933.295865 & 27.24 & 22.30 & 20.04 & 17.42 & 16.41 & 15.79 & 15.69 & 15.52 & 15.88 & l / early - t & + 7 & 218.001634 & @xmath8933.925557 & & 26.00 & 22.33 & 19.29 & 18.63 & 17.70 & 17.22 & 16.32 & & late - t & red galaxy ? + 8 & 218.003005 & @xmath8933.949375 & & 23.89 & 21.35 & 18.89 & 18.80 & 17.79 & 17.46 & 16.90 & 16.07 & late - t & red galaxy ? + 9 & 218.303091 & @xmath8934.477201 & 26.43 & 21.69 & 19.23 & 16.33 & 15.20 & 14.58 & 14.69 & 14.47 & 14.42 &
l / early - t & + 10 & 218.335896 & @xmath8933.850797 & & & & 20.25 & 18.86 & 17.12 & 17.01 & 16.40 & 15.21 & l / early - t & bad i photometry ?
+ -nn distance threshold , for @xmath72 , 5 and 7 ( using all colors or irac only colors ) . the prediction is based on a monte carlo simulation seeded with a 20% subsample of the first look survey database .
the large dots are the actual fractions of sources rejected from the full datasets for different values of @xmath1 . ] , @xmath59 and irac colors , satisfying the optical color requirement .
circles are l / early - t candidates and squares late - t candidates .
large symbols have been verified visually in the sdss or ndwfs i band images to ensure they are single point sources and not extended sources or blends .
small symbols are not detected in the optical surveys .
grey symbols are the brown dwarf templates .
the solid line in the bottom panel shows the @xmath0-nn region drawn for @xmath72 and @xmath94 using only the two variables in the plot ( @xmath92 $ ] and @xmath40 - [ 4.5]$ ] ) . ] | we present a statistical method for the photometric search of rare astronomical sources based on the weighted @xmath0-nn method .
a metric is defined in a multi - dimensional color - magnitude space based only on the photometric properties of template sources and the photometric uncertainties of both templates and data , without the need to define ad - hoc color and magnitude cuts which could bias the search .
the metric is defined as a function of two parameters , the number of neighbors @xmath0 and a threshold distance @xmath1 that can be optimized for maximum selection efficiency and completeness .
we apply the method to the search of l and t dwarfs in the _ spitzer _ extragalactic first look survey and the botes field of the _ spitzer _ shallow survey , as well as to the search of sub - stellar mass companions around nearby stars . with high level of completeness ,
we confirm the absence of late - t dwarfs detected in at least two bands in the first look survey , and only one in the shallow survey ( previously discovered by @xcite ) .
this result is in agreement with the expected statistics for late - t dwarfs .
one l / early - t candidate is found in the first look survey , and 3 in the shallow surveys , currently undergoing follow - up spectroscopic verification . finally , we discuss the potential for brown dwarf searches with this method in the _ spitzer _ warm mission exploration science programs . |
the spectra of high - redshift quasars suggest that cosmic reionization was completed by @xmath9 @xcite , and recent measurements of the kinetic sunyaev - zeldovich effect constrain the duration of this process to @xmath10 @xcite .
the galaxy population at @xmath1 may in principle be sufficient to reionize the universe ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , but this hinges on the slope of the galaxy luminosity function at luminosities significantly below current detection thresholds , and on the fraction of hydrogen - ionizing photons that escape from galaxies into the intergalactic medium ( igm )
. the latter quantity , the lyman - continuum ( lyc ) escape fraction @xmath0 , can be directly measured at @xmath11 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
. observations of this type indicate an increase in the typical @xmath0 with redshift in the @xmath123 interval @xcite , and data - constrained models for galaxy - dominated reionization also require a redshift evolution in @xmath0 .
simulations and theoretical arguments moreover suggest that @xmath0 may vary as a function of galaxy mass and star - formation activity . at redshifts @xmath13 ,
measurements of the rest frame lyc flux ( at wavelengths @xmath14 ) are precluded by the opacity of the increasingly neutral igm . to probe @xmath0 throughout the reionization epoch ,
indirect measurements are instead required .
some constraints on the _ typical _ @xmath0 in the reionization epoch may be obtained from fluctuations in the cosmic infrared background fluctuations @xcite , or by combining the observed galaxy luminosity function with ly@xmath15 forest data ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , but it remains unclear whether and how @xmath0 may be assessed for individual galaxies at these redshifts . here , we argue that since @xmath0 regulates the impact of nebular emission on the spectral energy distribution ( seds ) of galaxies , it should be possible to identify high-@xmath0 candidates from their rest - frame ultraviolet / optical seds at @xmath1 , at least for objects with very high escape fractions ( @xmath16 ) .
@xcite , @xcite and @xcite have previously attempted to assess @xmath0 from photometric data , but spectroscopic signatures are likely to produce more reliable results for individual targets .
recently , @xcite presented a spectroscopic method to place upper limits on the lyc escape fraction of high - redshift galaxies using metal absorption lines , but unlike their method , the one proposed here should be able to place both upper and lower limits on the escape fraction . in sect .
[ mechanisms ] , we describe the two main mechanisms capable of producing lyc leakage in galaxies and introduce the geometries adopted in our subsequent modelling of these . in sect .
[ signatures ] , we present a simple spectral diagnostic that may be used to assess @xmath0 in the case of high - leakage objects , and discuss the role of metallicity , star formation history and dust attenuation on the relevant spectral features . focusing on the capabilities of the _ near infrared spectrograph ( nirspec ) _ on the upcoming _ james webb space telescope ( jwst ) _
, we derive the galaxy luminosity limits for the proposed method to estimate @xmath0 in sect . [ discussion ] .
a number of lingering problems with the proposed method are also discussed .
[ summary ] summarizes our findings .
all our calculations are based on a @xmath17 , @xmath18 , @xmath19 km s@xmath20 mpc@xmath20 cosmology , in rough agreement with the constraints set by the wmap 9-year data combined with @xmath21 measurments and baryon acoustic oscillations @xcite and the first planck results @xcite .
whenever uv slopes and line equivalent widths are discussed , these quantities are given for rest - frame seds .
there are basically two different mechanisms that can cause lyc leakage from star - forming regions an radiation - bounded nebula with holes , and a density - bounded nebula ( also known as a truncated strmgren sphere ) .
these two scenarios are schematically illustrated in fig .
[ geometries ] , in which our model galaxy is depicted as a centrally concentrated ensemble of stars surrounded by a single hii region . the first case ( fig .
[ geometries]a ) corresponds to the situation when supernovae or stellar winds have opened up low - density channels in the neutral interstellar medium ( ism ) through which lyc photons may escape without getting absorbed . the second case ( fig .
[ geometries]b ) corresponds to a situation when the lyc flux from a very powerful star - formation episode `` exhausts '' all the hi before a complete strmgren sphere can form , thereby allowing lyc photons to escape into the igm .
galaxies are admittedly more complex than the simple , toy - model geometries depicted in fig . [ geometries ] and
contain a spatially extended ensemble of hii regions with different sizes and densities .
actual cases of lyc leakage are therefore likely to be due to mixtures of the two mechanisms . however , as we will demonstrate in sect .
[ ew_beta ] , the spectral diagnostics we propose are very similar for these two limiting cases , as long as dust effects can be ignored .
dust attenuation does , however , affect these two geometries differently , as discussed in sect .
[ attenuation ] .
the situation where lyc leakage is caused by high - velocity , lyc - emitting stars that venture far from the centres of galaxies is admittedly not well - represented by these simple geometries , but should effectively give rise to a situation similar to that in fig .
[ geometries]b .
the sed model results presented in this paper are based on the _ yggdrasil _ spectral synthesis code @xcite , which mixes single - age stellar population spectra to simulate arbitrary star formation histories .
these are used as input to the photoionization code cloudy @xcite .
the result is an age sequence of model seds which both stellar continuum emission , nebular continuum emission and nebular emission lines . in this paper , we adopt starburst99 stellar seds generated using padova - agb stellar evolutionary tracks at metallicities @xmath220.020 , assuming the @xcite universal stellar initial mass function ( imf ) throughout the mass range 0.1100 @xmath23 .
since we are focusing on galaxies well into the reionization epoch , we limit the discussion to objects at @xmath24 , with ages @xmath25 ( roughly the age of the universe at @xmath26 ) .
population iii galaxies ( at zero or near - zero metallicities ) may also exist at high redshifts and are predicted to exhibit pronouned lyc leakage @xcite . while such objects likely have too low masses to contribute substantially to the reionization of the universe , we briefly discuss models for such galaxies in sect .
[ popiii ] , using stellar seds from @xcite .
yggdrasil assumes that the nebular component of the overall galaxy sed can be treated as originating from a single , spherical and isotropic hii region ( as schematically depicted in fig .
[ geometries ] ) with constant density and filling factor . in the case of radiation - bounded nebulae with holes ( fig .
[ geometries]a ) , model seds with different @xmath0 are generated by varying the covering factor of the nebula in cloudy . in the case of density - bounded nebulae ( fig .
[ geometries]b ) , different @xmath0 are instead achieved by truncating the nebula at a fixed fraction of the theoretical strmgren radius . in this paper
, we adopt a constant hydrogen number density @xmath27 @xmath28 , a filling factor of 0.01 and a gas metallicity identical to that of the stars
. however , the primary diagnostics that we discuss ( balmer recombination lines and nebular continuum in the rest - frame uv / optical ) are not sensitive to these parameter choices .
we have moreover checked that our predictions for radiation - bounded nebulae with holes ( fig .
[ geometries]a ) are in reasonable agreement with those of the @xcite model , which is based on a slightly different computational machinery .
some of the ionizing photons produced by massive , hot stars are absorbed by the neutral hydrogen in the ism and produce prominent hii regions . as a result , photons at rest - frame wavelengths @xmath29 get reprocessed into emission lines and nebular continuum flux at longer wavelengths .
a substantial fraction of the observed rest - frame uv / optical fluxes of young and/or star - forming galaxies is therefore expected to come from nebular emission rather than direct star light . since the relative contribution from nebular emission to the overall galaxy spectrum becomes smaller if some fraction of the ionizing photons escape directly into the igm , information about the lyc escape fraction is imprinted in the non - ionizing ( @xmath30 ) part of the sed .
this is demonstrated in fig .
[ spectra ] , where we present model seds for young starburst galaxies ( age 10 myr ) at metallicities @xmath31 ( fig .
[ spectra]a ) and @xmath22 ( fig . [ spectra]b ) .
the line colours in each panel represent different @xmath0 ( 0.0 , 0.3 , 0.5 and 0.7 ) . at both metallicities ,
lower @xmath0 imply stronger emission lines and a more prominent balmer jump at 3646 . in fig .
[ spectra]b , the entire slope of the uv continuum changes with @xmath0 , beacuse of the higher relative contribution from nebular emission at low metallicity .
similar changes in the uv slope are seen at higher metallicities as well , albeit at younger starburst ages .
the noticeable dependence of several spectral features on @xmath0 suggest that it should be possible to assess the lyc escape fraction from galaxy seds without actually measuring the lyc flux .
the question is simply how to best retrieve the @xmath0 information from the rest - frame uv / optical sed that can actually be observed in the reionization epoch ( @xmath32 , since the flux at shorter wavelengths is absorbed by the neutral igm ) .
the power - law slope @xmath2 ( @xmath33 ) of the uv continuum @xmath34 can in principle be used to gauge @xmath0 , since the addition of nebular emission to a young , purely stellar sed tends to shift the slope of the uv continuum in the redward direction and increase @xmath2 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the interpretation becomes ambiguous , however , unless a very blue uv slope is detected ( @xmath35 ) .
the distribution of @xmath2 slopes among galaxies in the reionization epoch is still a matter of debate .
@xcite reported an average @xmath36 for @xmath7 , but later studies have given no support for a typical slope this extreme and instead indicate @xmath37 .
it is possible , however , that a subset of objects still display uv slopes substantially bluer than this ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the problem in inferring @xmath0 from @xmath2 is that @xmath2 also depends on age , metallicity and dust reddening .
this is demonstrated in fig .
[ uv_slope ] , where we plot the age dependence of @xmath2 for @xmath31 ( fig .
[ uv_slope]a ) and @xmath22 ( fig . [ uv_slope]b ) galaxies with either single - age stellar populations ( dashed lines ) or constant star formation rates ( solid lines ) .
different line colours are used to indicate the impact of the lyc escape fraction in the case of an radiation - bounded nebula with holes ( fig .
[ geometries]a ) .
the arrow shows the shift in @xmath2 ( @xmath38 ) predicted for @xmath39 mag of @xmath40-band extinction of the stellar continuum ( corresponds to @xmath41 mag at a rest wavelength of 1500 ) in the case of @xmath42 , assuming the @xcite average large magellanic cloud ( lmc ) extinction curve .
this amount of uv / optical extinction is consistent with several current estimates of the amount of dust affecting the uv slopes of @xmath24 galaxies ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . at fixed @xmath39 mag
, the @xcite attenuation curve gives a similar result ( @xmath38 ) .
as seen in fig .
[ uv_slope ] , the effect of @xmath0 on @xmath2 is very small for metal - rich galaxies ( fig .
[ uv_slope]a ) but becomes substantial at low metallicities ( fig .
[ uv_slope]b ) due to greater importance of nebular emission in the latter case .
but even at metallicities as low as @xmath22 , it is only for extremely blue uv slopes ( @xmath43 ) that any sort of constraint on @xmath0 can be set from @xmath2 alone . here
, we use the original definition of @xmath2 given by @xcite , in which the uv slope is derived from the overall continuum flux ( stellar and nebular ) in 10 wavelength intervals ( chosen to avoid stellar and interstellar absorption features ) in the range @xmath442580 .
the exact value of @xmath2 at any given age depends a bit on the exact wavelength region over which this slope is measured @xcite , but the overall trends seen in fig .
[ uv_slope ] remain the same , regardless of the definition of @xmath2 used .
it should be noted , however , that the 2175 dust feature can interfere with a few of the @xcite bands and make the uv slope behave in unexpected ways . in this paper , we therefore refrain from discussing attenuation recipies with very prominent 2175 bumps , like the average milky way extinction curve ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ? * ) .
we suggest that by combining the uv slope @xmath2 with the equivalent width ( ew ) of a balmer line such as h@xmath2 , it should be possible to identify galaxies with high escape fractions ( @xmath4 ) as long as the uv slope is @xmath45 after dust reddening corrections ( not far from the typical slope measured at @xmath1 , which is likely to be at least slightly affected by dust ; @xcite ) .
this idea is demonstrated in fig .
[ hb_beta_geom ] , where we plot the rest - frame ew(h@xmath2 ) against @xmath2 for @xmath31 for constant - sfr , dust - free models with different lyc escape fractions .
results are shown for both geometries depicted in fig .
[ geometries ] , i.e. an radiation - bounded nebula with holes ( solid lines ) and a density - bounded nebula ( dashed lines ) .
the high - leakage models ( @xmath4 ; orange , green and blue lines ) are all shifted to the right in fig .
[ hb_beta_geom ] ( i.e. towards lower ew(h@xmath2 ) ) compared to the no - leakage case ( red lines ) . since the model tracks remain separated up to the highest age considered ( 700 myr ; approximately the age of the universe at @xmath26 )
, this diagram suggests that @xmath0 can be assessed from a simultaneous measurement of ew(h@xmath2 ) and @xmath2 ( additional diagnostics can however improve the quality of this estimate , as discussed below ) . the results for radiation - bounded and the density - bounded models are next to identical at @xmath46 ( and , by definition , completely identical for @xmath42 ) . in the following ,
we will therefore focus on the radiation - bounded models . at this metallicity ( @xmath31 )
, @xmath2 is seen to change very little as @xmath0 is increased , simply due to the fact that nebular emission has such a small impact on the uv slope ( fig .
[ uv_slope ] ) .
the equivalent width ew(h@xmath2 ) of course drops with increasing @xmath0 , but since nebular emission produces both the h@xmath2 emission line and part of the continuum below the line , the decrease is not as dramatic as one would expect if one assumed the underlying continuum to be purely stellar .
similar diagnostic diagrams can in principle be formed for other balmer lines ( h@xmath15 , h@xmath47 , h@xmath48 etc . ) , but h@xmath2 seems to be a good compromise between observational limitations ( h@xmath15 redshifts out of the jwst / nirspec range already at @xmath49 ) and expected diagnostic value ( balmer lines beyond @xmath3 are weaker and have lower equivalent widths ) .
other definitions of @xmath2 give rise to slightly different diagnostic diagrams , but the overall trends remain the same . while the [ oiii]@xmath505007 emission line can be measured over approximately the same redshift interval as h@xmath2 , and will in many cases be stronger , the equivalent width of this line is not as suitable as a diagnostic , since it is very sensitive to the ionization parameter of the gas .
since metallicity is an important parameter in regulating the impact of nebular emission in galaxies @xcite , one expects diagnostic diagrams such as fig .
[ hb_beta_geom ] to show some metallicity dependence . in fig .
[ hb_beta_z ] , we display the same @xmath31 , constant - sfr model tracks for radiation - bounded nebulae as in fig .
[ hb_beta_geom ] , alongside otherwise identical models at @xmath51 and @xmath22 .
as expected , @xmath2 evolves more strongly with @xmath0 at low metallicities
( see fig .
[ uv_slope ] ) , whereas ew(h@xmath2 ) displays a milder metallicity dependence .
the latter effect is due to the greater contribution from nebular emission to _ both _ the h@xmath2 emission line and the underlying continuum at low metallicities .
while a rough assessment of @xmath0 can be obtained even if the metallicity is unknown , a more precise estimate requires some handle on the metallicity .
this is most readily obtained from the strengths of emission lines like [ neiii]@xmath503869 , [ oii]@xmath503727 and [ oiii]@xmath504959 , 5007 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , since data on these lines come `` for free '' when doing spectroscopy over the rest - frame wavelength interval ( @xmath525000 ) required to measure @xmath2 and ew(h@xmath2 )
. it should be noted , however , that many of the metallicity calibrations developed for these lines are based on the assumption of no lyc leakage , and may potentially give biased results if blindly applied to density - bounded nebulae with high @xmath0 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
as illustrated in fig .
[ hb_beta_sfh ] , temporal variations of the star formation rate ( sfr ) can affect both ew(h@xmath2 ) and @xmath2 and need to be considered when attempting to assess the lyc escape fraction .
galaxies with increasing star formation rates simply remain longer in the high ew(h@xmath2 ) , blue @xmath2 part of the ew(h@xmath2)-@xmath2 diagram , which makes the determination of @xmath0 easier than in the case of a constant sfr ( fig .
[ hb_beta_geom ] ) .
this is shown in fig .
[ hb_beta_sfh]a , where @xmath31 models with exponentially increasing sfrs ( @xmath53 with @xmath54 yr ) at various @xmath0 are used .
models with decreasing star formation rates are , however , more troublesome . in fig .
[ hb_beta_sfh]b , we show the behaviour of models with exponentially declining star formation rates ( @xmath55 yr ) .
these models evolve faster than the constant - sfr models of fig .
[ hb_beta_geom ] and reach slightly redder @xmath2 at the highest ages considered .
however , as there is no overlap between the different @xmath0 tracks , it should still be possible to estimate the lyc escape fraction . in fig .
[ hb_beta_sfh]c , we consider a more extreme model with a very short burst ( @xmath56 yr ) of constant star formation , after which the sfr immediately drops to zero .
since the lyman continuum flux decreases very quickly once star formation has ceased , models of this type evolve sharply to the right in the ew(h@xmath2)-@xmath2 diagram once the burst is over . during the post - burst phase
, there is a brief period where model lines corresponding to different @xmath0 overlap , which means that an accurate assessment of @xmath0 would be very difficult in this part of the diagram . for these particular models ,
this happens at ages @xmath5715 myr ( @xmath58 ) , when the lyc flux has dropped to 215% of what it was right before star formation ceased .
we stress , however , that galaxies will only rarely be caught in this brief transition phase .
single - component models of the type used in fig .
[ hb_beta_sfh]a c are often assumed to be adequate in the interpretation of observational data on high - redshift galaxies , but real galaxies may well exhibit more complicated star formation histories , with several bursts with intermittent periods of low star formation in between . in fig . [ hb_beta_sfh]d , we consider a sequence of equal - strength , @xmath56 yr long , constant - sfr bursts once every @xmath59 yr , with zero sfr in between .
each burst shifts the models to slightly redder @xmath2 and lower ew(h@xmath2 ) , thereby causing a slight overlap between models with different @xmath0 . in this context
, the worst - case scenario would be to have a single young , low-@xmath0 burst superposed on an old , passively evolving population which contributes substantially to the h@xmath2 continuum ( thereby lowering ew(h@xmath2 ) ) while leaving the uv slope @xmath2 largely unaffected , since this could in principle mimic high @xmath0 in the ew(h@xmath2)-@xmath2 diagram . in fig .
[ hb_beta_sfh2]a , we show the evolution of such a composite population , where a 1 myr old , @xmath42 model has been superposed on a 700 myr old , passively evolving population . as the mass ratio between the old population and the short burst
is increased to @xmath60 , this @xmath42 model can wander into the region normally occupied by leaking galaxies with @xmath16 .
if the mass ratio is increased to @xmath61 , the passively evolving population becomes completely dominant also in the uv , and shifts @xmath2 to very red values .
however , the presence of an old , massive component gives rise to other tell - tale spectral signatures at the wavelengths probed by both nirspec ( 0.65 @xmath62 m ) and the mid - infrared instrument ( miri ; 527 @xmath62 m in imaging mode ) onboard the jwst .
this is demonstrated in fig .
[ hb_beta_sfh2]b , where we plot the seds of a 10 myr old , constant - sfr model with @xmath63 ( blue line ) and a 1 myr old @xmath42 model superposed on a passively evolving population with age 700 myr ( black line ) , where the old component is 300 times massive than the young population .
both spectra have been redshifted to @xmath26 .
these two models occupy a very similar position in the ew(h@xmath2)-@xmath2 diagram ( black triangle and orange @xmath56-yr age marker in fig .
[ hb_beta_sfh2]a ) , but the continuum flux observed at @xmath64 m would be very different in the two cases ( much higher in the case of the two - component galaxy ) .
the continuum flux at @xmath65 m comes for free when measuring @xmath2 and ew(h@xmath2 ) with jwst / nirspec , and jwst / miri can trace the continuum to even longer wavelengths .
miri lacks the sensitivity to provide useful spectroscopic limits on the continuum , but miri imaging should yield competitive constraints .
this is demonstrated in fig .
[ hb_beta_sfh2]b , where we include the expected detection limits for jwst / nirspec low - resolution ( @xmath66 ) , 15 @xmath62 m spectroscopy ( red dashed line ) and compare these to the corresponding limits for jwt / miri imaging ( red triangles ) at @xmath6710 @xmath62 m . in both cases ,
we adopt an exposure time of 10h and require a @xmath68 detection . in the quest to derive @xmath0 from the rest - frame uv / optical sed
, dust attenuation is potentially a more serious problem than metallicity or star formation history , since this part of the sed may contain insufficient information to gauge the impact of lyc extinction .
many recent studies of reionization - epoch galaxies suggest that these objects have suffered very little dust attenuation ( @xmath69 mag ; e.g. @xcite , but see for a different view ) . on the other hand , it does nt necessarily take much attenuation in the rest - frame optical or non - ionizing uv to have a considerable impact on the lyc it depends on the spatial distribution of the dust compared to the gas and stars .
predicting the effects of dust on galaxy seds is notoriously difficult , since the effective dust attenuation law depends not only on the amount and composition of dust , but also on the spatial distribution of stars dust and gas . in fig .
[ dust_geometry ] , we schematically illustrate four different dust distributions ( two for each of the gas geometries of fig .
[ geometries ] ) .
this is not meant to be an exhaustive list of the possible dust distributions , but serves to illustrate a few different scenarios . in fig .
[ dust_geometry]a and b , the dust is distributed in a uniform screen _ outside _ the hii region . in the cases depicted in fig .
[ dust_geometry]a and c ( ioniziation - bounded nebulae with holes ) , dust has no direct impact on the leaking lyc flux , since the leakage is assumed to take place through holes that are devoid of both gas and dust . in the density - bounded nebula of fig .
[ dust_geometry]b , on the other hand , the dust screen has a very pronounced effect on the escaping lyc . as an example of this , consider a dust screen giving a rest - frame @xmath40-band extinction of @xmath70 mag .
extinction curves are notoriously uncertain in the far - uv , but if one adopts the average lmc attenuation curve presented by @xcite , this @xmath71 converts into a predicted lyc depletion factor of @xmath72 at 910 , but the @xcite milky way extinction curve which extends to shorter wavelengths predicts a similar lyc depletion factor ] .
hence , the lyc escape fraction would effectively be limited to @xmath73 . in the case of a foreground dust screen ( fig .
[ dust_geometry]a and b ) , the rest - frame optical attenuation can straightforwardly be estimated from jwst / nirspec spectra of @xmath99 galaxies by measuring of the ratios of balmer emission - lines ( e.g. h@xmath2/h@xmath47 ) , since the intrinsic ratio can be estimated from recombination theory ( h@xmath2/[email protected] for a wide range of electron densities and temperatures , assuming case b recombination ) .
however , if the dust is spatially mixed with the hii gas as in fig .
[ dust_geometry]c and d , some fraction of the lyc photons will be absorbed by dust before they have a chance to ionize the gas ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and this will not be revealed by the balmer line ratios . while radiation pressure and dust sublimation act to produce a dust cavity around the ionizing stars , neither of these effects is sufficiently strong to prevent lyc photons from getting absorbed directly by dust in local hii regions @xcite . indeed ,
@xcite estimate that @xmath75 of the lyc photons produced in low - redshift starburst galaxies are typically absorbed by dust .
since simulations of high - redshift galaxies suggest that when lyc escapes , it does so through essentially dust - free channels , we will limit the discussion to radiation - bounded nebulae with holes ( fig .
[ dust_geometry]a and c ) . for both of these geometries ,
the effects of small amounts of optical attenuation ( @xmath70 mag ) on positions of galaxies in the ew(h@xmath2)-@xmath2 diagram are illustrated in fig .. [ dust_corrections ] , assuming the @xcite average lmc attenuation law to capture the transport of radiation through dusty regions . at low @xmath0 , a foreground dust screen ( attenuation vector represented by solid arrow )
has a pronounced effect on @xmath2 but almost no effect on ew(h@xmath2 ) , since the stellar continuum emerging through the screen still dominates the rest - frame uv , while the unattenuated continuum that escapes through the holes remains subdominant at wavelengths close to h@xmath2 . at high @xmath0 ,
the reverse is true : @xmath2 is largely unaffected by dust since the unattenuated stellar uv continuum escaping through the holes dominates over the dust - attenuated continuum , and ew(h@xmath2 ) is lowered due to the increasing importance of stellar h@xmath2 continuum emerging through the holes .
hence , the attenuation vector changes direction and amplitude depending on where in the ew(h@xmath2)-@xmath2 diagram one started off . if the dust is spatially mixed with the ionized gas ( fig .
[ dust_geometry]c ) , the stellar non - ionizing continuum will still suffer @xmath70 mag of attenuation , but the balmer line ratios will reflect a value smaller than this , since part of the attenuation happened prior to photoionization . at fixed @xmath0 , the nebula will also appear fainter due to lyc extinction . here , we model this case by assuming that the star light entering the hii region is attenuated by @xmath76 mag before it has a chance to ionize the gas ( lowering the number of lyc photons that enter the gas by a factor of @xmath77 ) , and apply a further @xmath76 mag correction outside the hii region ( giving a total of @xmath70 mag , as before ) . other ratios of stellar to nebular attenuation can certainly be considered , but this serves to exemplify the ew(h@xmath2 ) and @xmath2 trends . the resulting lyc depletion factor is somewhat lower than suggested by @xcite and @xcite for local hii regions , but similar to that derived by @xcite for low - redshift starburst galaxies .
our procedure gives rise to the dashed attenuation arrows in fig .
[ dust_corrections ] . as in the case of the outer dust screen ,
the @xmath2 slope is reddened more at low @xmath0 than at high @xmath0 , but the overall reduction in nebular emission ( due to lyc extinction ) increases the relative impact of stellar h@xmath2 continuum on the emerging sed and decreases ew(h@xmath2 ) substantially at all @xmath0 . at fixed @xmath78 ,
the two types of attenuation vectors discussed above have slightly different directions and amplitudes .
moreover , a measurement of the h@xmath2/h@xmath47 balmer decrement will tend to underestimate the overall attenuation whenever lyc photons are lost due to dust extinction , as in the case where ionized gas and dust are mixed . for a given h@xmath2/h@xmath47 ratio , there will therefore be lingering uncertainties in how to correct the position in the ew(h@xmath2)-@xmath2 diagram unless the amount of lyc extinction can be assessed .
for example , consider an object observed to lie close to the @xmath56 yr marker on the orange line ( @xmath63 ) in fig .
[ dust_corrections ] . to correct the position of this object for dust effects
, one needs to pick the right attenuation vector , and to move the object in the opposite direction in the diagram .
if the balmer line ratios suggest @xmath70 mag , a correction assuming an outer dust screen would place the object on the green line ( implying @xmath79 ) , whereas a correction assuming mixed gas and dust would still place it on the orange line ( i.e. @xmath63 ) .
these uncertainties are not easy to overcome from observations in the rest - frame uv / optical alone , but supporting observations in the rest - frame mid / far - ir could improve the situation ( as further discussed in sect .
[ ir ] ) .
the possibility that the effective attenuation law may have a more pronounced effect on emission lines than on the far - uv / optical continuum represents yet another complication .
this phenomenon is observed in local starburst galaxies ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and is most readily interpreted as due to temporal evolution in the opacity of star - forming regions the youngest star clusters ( that contribute most of the lyc photons and the nebular emission ) tend to be more strongly affected by dust than the slightly older star clusters , which contribute substantially to the far - uv / optical stellar continuum ) . in principle , this effect could mimic lyc leakage in the ew(h@xmath2)-@xmath2 diagram by causing a substantial reduction in ew(h@xmath2 ) but only a modest reddening of @xmath2 .
however , we find that the @xcite attenuation scheme in which the nebular emission experiences a dust optical depth a factor of @xmath77 times than that of the stars results in an attenuation vector no more extreme than those already included in fig . [ dust_corrections ] . in order for a @xmath42 galaxy to be misclassified as an @xmath4 object , one requires both a higher ratio of nebular to stellar dust opacity as in some of the more extreme objects discussed by @xcite and substantial nebular attenuation ( @xmath80 mag , as revealed by e.g. the h@xmath81h@xmath47 emission line ratio ) .
the radiation absorbed by dust is reradiated in the rest - frame mid / far - ir , and combined observations of the rest - frame uv / optical and mid / far - ir have been used to estimate the fraction of lyc photons absorbed by dust in local objects ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this dust emission lies out of range of the jwst for @xmath1 objects , but as further discussed in sect .
[ detection_limits ] , observations with the atacama large millimeter / submillimeter array ( alma ) or the planned space infrared telescope for cosmology and astrophysics ( spica ) may be able to probe dust emission from galaxies at these redshifts .
there are , however , many outstanding problems involved . for instance , the dust is heated by both the lyc , the non - ionizing uv and the resonant ly@xmath15 line . depending on the spatial distribution of gas , dust and stars and the overall velocity field
, ly@xmath15 photons may escape from starburst galaxies at all redshifts , but for most galaxies in the reionization epoch , a substantial fraction of the ly@xmath15 photons that escape are scattered into huge and largely unobservable ly@xmath15 halos by the neutral igm ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , thereby causing potentially significant uncertainties in the ir radiation budget .
the sed of dust emission also depends on the dust composition @xcite , which is highly uncertain even for low - redshift targets .
the part of the sed required for the measurement of @xmath2 and ew(h@xmath2 ) falls within the wavelength range of the jwst / nirspec spectrograph for @xmath829 . the use of higher - order balmer lines instead of h@xmath2 ( h@xmath47 , h@xmath48 etc . )
would extend the range to @xmath83 , but at the expense of losing the [ oiii]@xmath505007 line , which is useful for getting a handle on the metallicity ( sect .
[ metallicity ] ) .
since the proposed method is applicable for relatively blue seds ( @xmath84 ) and positive ew(h@xmath2 ) , the most challenging aspect of these observations is to measure the @xmath3 continuum .
using the jwst / nirspec exposure time calculator version p1.5 with the @xmath66 setting , we estimate that one should be able to get a @xmath85 measurement of the h@xmath2 continuum ( giving an error @xmath86 , since the error on the line flux will be negligible by comparison ) by integrating over @xmath87 bins in the rest - frame spectrum ) in 10 h for a 100 myr old , @xmath31 , @xmath63 , constant - sfr galaxy with stellar mass @xmath88 at @xmath26 . at @xmath89 ,
the corresponding mass limit would be @xmath90 .
galaxies this massive are well within the detection limits of ultradeep hubble space telescope imaging , and many @xmath91@xmath92 objects have already been discovered in current @xmath24 samples .
strong lensing by foreground galaxy clusters may boost the fluxes of high - redshift background objects by magnification factors of up to @xmath93100 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . by targeting lensed fields
, it should therefore be possible to push these detection limits down to @xmath94 ( @xmath26 ) and @xmath95 ( @xmath89 ) , if @xmath93 is assumed . in terms of the rest - frame 1500
luminosities often used when measuring the luminosity function at high redshifts , this corresponds to @xmath6 at @xmath26 and @xmath8 at @xmath89 .
these detection limits convert into star - formation rates of @xmath96 , placing them slightly above the level where stochastic sampling of the stellar initial mass function starts to have serious effects on the lyc flux of galaxies . on a somewhat longer timescale , similar observations will also be possible with groundbased telescopes like the _ giant magellan telescope _ , the _ thirty meter telescope _ and the _
european extremely large telescope_. as discussed in sect .
[ attenuation ] , complementary observations of the rest - frame mid / far - ir with spica or alma could in principle provide information on the fraction of lyc photons destroyed directly by dust .
such observations could also reveal the presence of heavily embedded components within the target galaxies .
formally , the @xmath0 derived from jwst observations can only reflect the star formation that actually contributes to the rest - frame uv / optical sed , and extremely dust - obscured star formation ( undetectable in the optical , but bright at ir wavelengths , as in infrared - luminous galaxies at lower redshifts ; e.g. @xcite ) would shift this @xmath0 estimate away from that relevant for simulations of galaxy - driven reionization , where the _ total _ star formation rate of galaxies needs to be converted into a lyc flux emitted into the igm . based on the dust
seds for star - forming dwarf galaxies presented by @xcite , we estimate that strongly lensed ( @xmath97 ) , @xmath99 galaxies with sfr @xmath98 yr@xmath20 should be detectable with alma in @xmath99 h exposures at @xmath100@xmath101 m .
spica may be able to detect such lensed @xmath9 galaxies at @xmath102 m in less than 1 h , but this requires that one can push the photometry a factor @xmath67 below the formal spica confusion limit for instance by using auxiliary , high - resolution data from other wavelength bands ( e.g. obtained with jwst or alma ) to subtract off the flux contribution from nearby interlopers . in its current form
, the proposed method can be used to assess @xmath0 for the brightest galaxies at @xmath1 , but is likely to provide interesting limits for galaxies exhibting very high lyc escape fractions only ( @xmath16 ) .
however , the rest - frame uv / optical sed carries more information than contained in the ew(h@xmath2)-@xmath2 diagnostic , and stronger constraints can possibly be set by making use of the entire sed obtained from jwst / nirspec observations . in future papers , we intend to explore the true @xmath0 limits that can be obtained from spectroscopy using more detailed sed simulations . while it is likely that @xmath0 estimates for less extreme leakers may be obtained this way , the luminosity limits of the method are unlikely to change significantly .
hence , only objects with @xmath6 ( stellar masses @xmath103 ) can be probed this way . galaxies in this mass range may be insufficient to explain the ionization state of the universe at @xmath104 , and methods based on photometry rather than spectroscopy ( in the vein of @xcite , @xcite , @xcite ) will be necessary to estimate @xmath0 for objects at lower masses . while photometric methods are unlikely to provide strong constraints for individual targets , they can on the other hand
be applied to larger samples of galaxies . in the future ,
we therefore also intend to investigate the likely @xmath0 limits that can be set using photometry with jwst / nircam and jwst / miri filters .
population iii galaxies may exist at high redshifts and are predicted to have high @xmath0 @xcite .
such objects may be detectable with jwst in strongly lensed fields @xcite and should be identifiable because of their unusual colours @xcite and spectra ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , at least for ages up to @xmath105 yr @xcite .
while the low masses predicted for such galaxies ( total stellar mass perhaps no more than @xmath106 ; @xcite ) makes it unlikely that these objects would contribute substantially to the reionization of the universe , the spectral diagnostics we propose could in principle be used to assess @xmath0 for these galaxies as well .
this is demonstrated in fig .
[ popiii_fig ] , where we show the evolution predicted in the ew(h@xmath2)-@xmath2 diagram for a population iii galaxy experiencing a brief burst ( @xmath56 yr ) of zero - metallicity star formation , in the case of a radiation - bounded nebula with holes .
the characteristic masses of population iii star formation is generically predicted to be top - heavy , but the exact stellar imf remains highly uncertain .
here , we have used the @xcite stellar seds for a log - normal imf with characteristic mass @xmath107 , dispersion @xmath108 and wings extending from 1500 @xmath23 , but the results are similar for other imfs explored by @xcite during the early phase @xmath105 yr when population iii galaxies should be uniquely identifiable . at low ages ,
the strong impact of nebular emission on the uv slope at these low metallicities makes the model tracks curve upward in the ew(h@xmath2)-@xmath2 , but the tracks corresponding to different lyc escape fractions still remain clearly separated , thereby allowing an estimate of @xmath0 . at higher ages ( @xmath109 yr ) , ew(h@xmath2 ) drops while @xmath2 evolves very little , which causes substantial degeneracies between @xmath0 and the imf ( not shown , to avoid cluttering ) .
one of the most important problems in observational cosmology is identifying the sources that reionized the universe .
it is generally assumed that a substantial fraction of the ionizing photons were produced by galaxies , but other objects such as quasars ( e.g. * ? ? ?
* ) , mini - quasars ( e.g. * ? ? ?
* ) , micro - quasars ( e.g. * ? ? ?
* ) and population iii stars ( e.g. * ? ? ?
* ) may also have played a role .
currently , there exist mostly indirect constraints on the timing of the reionization epoch .
the spectra of high - redshifts quasars suggest that reionization was completed by @xmath9 @xcite , while observations of the cosmic microwave background can be used to put limits on the extent of the reionization period ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the detailed reionization history at @xmath110 remains highly uncertain , with statistics of ly@xmath15 emitters seemingly suggesting that the igm was still highly neutral at @xmath111@xmath112 ( e.g. @xcite ) , while igm temperature measurements instead indicate that reionization was already mostly complete by these redshifts @xcite .
the most promising prospect for constraining the reionization history in the near future is by observing the 21-cm line emission from the partly neutral igm .
radio observatories such as lofar ( van haarlem et al .
2013 , subm . )
, mwa @xcite and paper @xcite are aiming for a detection at some point within the next few years .
the 21-cm signal encodes a wealth of information about the physics of the reionization epoch , but the first generation of experiments will be focused on simple statistics such as the variance and power spectrum of the 21-cm brightness temperature .
these statistics will not be able to constrain the evolution of the ionized fraction directly , but will have to be interpreted by comparing to simulations ( e.g. @xcite ) .
most simulations to date include only galaxies as sources of ionizing photons , and use simplistic recipes to assign ionizing fluxes to the sources , tuned to reproduce the available observational constraints . with measurements of the lyc escape fraction in combination with
observed uv luminosity functions , much of the `` wiggle room '' for these recipes would be eliminated .
this in turn would make 21-cm measurements more powerful for constraining the sources of reionization by enabling better comparison between the 21-cm statistics predicted by different source models .
if galaxies were responsible for the reionziation of the universe , significant lyc photon escape from these objects must have occurred at @xmath1 . our ability to put this scenario to the test
is hampered by the neutral igm at these epochs , which precludes a direct measurement of the leaking lyc flux .
however , since the lyc escape fraction @xmath0 regulates the relative impact of nebular emission on the seds of galaxies , indirect information on this parameter should be retrievable from the sed . as a first demonstration of this
, we argue that extreme lyc leakers ( with @xmath16 ) may be identifiable from spectroscopic measurements of the equivalent width ew(h@xmath2 ) of the h@xmath2 emission line and the uv slope @xmath2 . by targeting strongly lensed galaxies , jwst /
nirspec spectroscopy should allow this technique to be applied to galaxies with stellar masses @xmath103 and redshifts up to @xmath5 .
we explore the impact of metallicity , star formation history and dust extinction on the ew(@xmath3 ) and @xmath2 diagnostics , and while all of these effects may complicate @xmath0 estimates to some extent , both the metallicity and the star formation history can be constrained by other spectral features that are readily detectable by jwst / nirspec .
dust represents a more troublesome issue , since the rest - frame uv / optical sed accessible to jwst / nirspec at @xmath99 may contain insufficient information to properly gauge this effect . while several studies have indicated that lyman - break galaxies in the reionization epoch have suffered very little dust reddening , direct absorption of lyc photons by dust may still occur , with lingering uncertainties in the inferred lyc escape fraction as a result .
heavily obscured star formation that contributes very little to the rest - frame uv / optical may in principle also bias @xmath0 estimates based on this part of the sed .
we suggest that supplementary observations of lensed @xmath99 galaxies with alma or the planned spica mission may help constrain such scenarios by measuring the dust emission peak in the rest - frame mid / far - ir . e.z acknowledges funding from the swedish national space board and the swedish research council .
a.k.i . acknowledges funding from the ministry of education , culture , sports , science , and technology ( mext ) of japan ( kakenhi : 23684010 ) .
the anonymous referee is acknowledged for helpful comments on the manuscript .
ade , p.a.r .
, et al . 2013 , arxiv1303.5076 alvarez , m. a. , finlator , k. , trenti , m. 2012 , apj , 759 , l38 benson , a. , venkatesan , a. , shull , j. m. 2013 , apj , 770 , 76 bergvall , n. , zackrisson , e. , andersson , b .-
g . , arnberg , d. , masegosa , j. , stlin , g 2006 , a&a , 448 , 513 bergvall , n. , leitet , e. , zackrisson , e. , marquart , t. 2013 , a&a , in press ( arxiv1303.1576 ) bogosavljevi , m. 2010 , phd thesis , california institute of technology boutsia , k. , et al .
2011 , apj , 736 , 41 bouwens , r. j. et al .
2010 , apj , 708 , l69 bouwens , r. j. , et al .
2012a , apj , 752 , l5 bouwens , r. j. , et al .
2012b , apj , 754 , 83 calzetti d. , kinney a.l .
, storchi - bergmann t. 1994 , apj , 429 , 582 calzetti d. , armus , l. , bohlin , r. c. , kinney a.l . , koornneef , j. , storchi - bergmann t. 2000 , apj , 533 , 682 cartledge s.i.b . , et al . 2005 , apj , 630 , 355 charlot , s. , & fall , s. m. 2000 , apj , 539 , 718 choi , p. i. , et al . 2006 , apj , 637 , 227 conroy , c. , kratter , k. m. 2012 , apj , 755 , 123 coe , d. et al . , 2013 , apj , 762 , 32 dijkstra , m. , haiman , z. , loeb a. 2004 , apj , 613 , 646 dunlop j.s .
2012 , arxiv1212.0860 fan , x. , et al .
2006 , aj , 132 , 117 ferland , g. j. , korista , k. t. , verner , d.a . , ferguson , j.w .
, kingdon , j.b . , verner , e.m .
1998 , pasp , 110 , 761 fernandez , e. r. , shull , j. m. 2011 , apj , 731 , 20 fernandez , e. r. , dole , h. , iliev , i. t. 2012 , arxiv1206.4934 ferrara , a. , & loeb , a. 2012 , arxiv1209.2123 finkelstein , s. l. , et al .
2012 , apj , 758 , 93 finkelstein s. l. , et al .
2012 , apj , 756 , 164 fontanot , f. , cristiani , s. , vanzella , e. 2012 , mnras , 425 , 1413 forero - romero , j. e. , dijkstra , m 2013 , mnras , 428 , 2163 giammanco , c. , beckman , j. e. , cedrs , b. 2005 , a&a , 438 , 599 gnedin , n. y. , kravtsov , a. v.,chen , h .- w .
2008 , apj , 672 , 765 gordon , k. d. , calzetti , d. , witt , a. n. 1997 , apj , 487 , 625 gordon k.d . ,
cartledge s. , clayton g.c . , 2009 , apj , 705 , 1320 grazian , a. , et al .
2012 , arxiv1208.0506 hinshaw , g. , et al .
2012 , arxiv1212.5226 hirashita , h. , buat , v. , inoue , a. k. , 2003 , a&a , 410 , 83 iliev , i. t. , mellema , g. , shapiro , p. r. , et al .
2012 , mnras , 423 , 222 inoue , a. k. , hirashita h. , kamaya , h. 2001 , apj , 555 , 613 inoue , a. k. , iwata , i. , deharveng , j .-
2006 , mnras , 371 , l1 inoue , a k. , iwata , i. 2008 , mnras , 387 , 1681 inoue , a. k. 2001 , aj , 122 , 1788 inoue , a. k. 2002 , apj , 570 , 688 inoue , a. k. 2010 , mnras , 401 , 1325 inoue , a. k. 2011 , mnras , 415 , 2920 iwata , i. , et al .
2009 apj , 692 , 1287 jeeson - daniel , a. , ciardi , b. , maio , u. , pierleoni , m. , dijkstra , m. , maselli , a. 2012 , mnras , 424 , 2193 jensen , h. , laursen , p. , mellema , g. , et al .
2013 , mnras , 428 , 1366 jiang , l. , et al .
2013 , arxiv1303.0024 johansson , d. et al . , 2012 , a&a , 543 , 62 johnson , j. l. , greif , t. h. , bromm , v. , klessen , r. s. , ippolito , j. 2009 , mnras , 399 , 37 jones , t. , ellis , r. s. , schenker , m. a. , stark , d. p. 2013
, arxiv1304.7015 lidz , a. , zahn , o. , mcquinn , m. , zaldarriaga , m. , & hernquist , l. 2008 , apj , 680 , 962 komatsu , e. , et al .
2011 , apjs , 192 , 18 kroupa , p. 2001 , mnras , 322 , 231 kuhlen , m. , & faucher - gigure , c .- a .
2012 , mnras , 423 , 862 larson , d. , dunkley , j. , hinshaw , g. , et al .
2011 , apjs , 192 , 16 leitet , e. , bergvall , n. , hayes , m. , linn , s. , zackrisson , e. 2013 , a&a , in press ( arxiv1302.6971 ) leitherer c. , et al . 1999 , apjs , 123 , 3 lorenzoni , s. , bunker , a. j. , wilkins , s. m. , stanway , e. r. , jarvis , m. j. , caruana , j. 2011 , mnras , 414 , 1455 madau , p. , rees , m.j . , volonteri , m. , haardt , f. , oh , s.p .
2004 , apj , 604 , 484 mclure , r. j. , dunlop , j. s. , cirasuolo , m. , koekemoer , a. m. , sabbi , e. , stark , d. p. , targett , t. a. , ellis , r. s. 2010 , mnras , 403 , 960 mclure , r. j. , et al .
2011 , mnras , 418 , 2074 mcquinn , m. , lidz , a. , zahn , o. , et al .
2007 , mnras , 377 , 1043 mirabel , i. f. , dijkstra , m. , laurent , p. , loeb , a. , pritchard , j. r. 2011 , a&a , 528 , 149 mitra , s. , ferrara , a. , choudhury , t. r. 2013 , mnras , 428 , l1 mortlock , d. j. , warren , s. j. , venemans , b. p. , et al .
2011 , nature , 474 , 616 nagao t. , maiolino r. , marconi a. 2006 , a&a , 459 , 85 nestor , d. b. , shapley , a. e. , steidel , c. c. , siana , b. 2011 , apj , 736 , 18 nestor , d. b. , shapley , a. e. , kornei , k. a. , steidel , c. c. , siana , b. 2013 , apj , 765 , 47 oesch , p. a. , et al .
2010 , apj , 709 , l16 ono , y. , ouchi , m. , shimasaku , k. , dunlop , j. , farrah , d. , mclure , r. , okamura , s. 2010 , apj , 724 , 1524 ouchi , m. , shimasaku , k. , furusawa , h. , et al .
2010 , apj , 723 , 869 paardekooper , j .-
p . , khochfar , s. , dalla , c. v. 2013 , mnras , 429 , l94 pandolfi , s. , ferrara , a. , choudhury , t. r. , melchiorri , a. , mitra , s. 2011 , phrvd , 84 , 123522 parsons a. r. et al . , 2010 ,
j , 139 , 1468 pentericci , l. , fontana , a. , vanzella , e. , et al .
2011 , apj , 743 , 132 pirzkal , n. , rothberg , b. , nilsson , k. k. , finkelstein , s. , koekemoer , a. , malhotra , s. , rhoads , j. 2012 , apj , 748 , 122 pirzkal , n. , rothberg , b. , ryan , r. , coe , d. , malhotra , s. , rhoads , j. , noeske , k. 2013 , arxiv1304.4594 raiter , a. , schaerer , d. , fosbury , r. a. e. 2010 a&a 523 , 64 raskutti , s. , bolton , j. s. , wyithe , j. s. b. , & becker , g. d. 2012 , arxiv e - prints 1201.5138 razoumov , a. o. , & sommer - larsen , j. 2010 , apj , 710 , 1239 robertson b.e . ,
ellis r.s . , dunlop j.s . ,
mclure r.j .
, stark d.p .
2010 , nature , 468 , 49 robertson b.e . ,
2013 , arxiv1301.1228 rogers , a. b. , mclure , r. j. , dunlop , j. s. 2013 , mnras , 429 , 2456 safranek - shrader c. , agarwal m. , federrath c. , dubey a. , milosavljevic m. , bromm v. 2012 , mnras , 426 , 1159 schaerer d. 2002 , a&a , 382 , 28 schaerer d. , de barros s. 2009 , a&a , 502 , 423 schaerer , d. , de barros , s. 2010 , a&a , 515 , 73 schaerer , d. , & pell , r. 2005 , mnras , 362 , 1054 siana , b. , et al .
2010 , apj , 723 , 241 steidel , c. c. , pettini , m. , adelberger , k. l. 2001 , apj , 546 , 665 shapley , a. e. , steidel , c. c. , pettini , m. , adelberger , k. l. , erb , d. k. 2006 , apj , 651 , 688 stiavelli , m. , trenti , m. 2010 , apj , 716 , l190 takeuchi , t. t. , ishii , t. t. , nozawa , t. , kozasa , t. , hirashita , h. 2005 , mnras , 362 , 592 theuns , t. , zaroubi , s. , kim , t .- s . , tzanavaris , p. , & carswell , r. f. 2002 , mnras , 332 , 367 tingay s. j. et al . , 2012 ,
arxiv e - prints 1206.6945 vanzella , e. , et al .
2010 , apj , 725 , 1011 vanzella , e. , et al .
2012 , apj , 751 , 70 vzquez , g. a. , & leitherer , c. 2005 , apj , 621 , 695 venkatesan , a. , tumlinson , j. , shull , j.m .
2003 , apj , 584 , 621 wild , v. , charlot , s. , brinchmann , j. , heckman , t. , vince , o. , pacifici , c. , chevallard , j. 2011 , mnras , 417 , 1760 wilkins s.m . , bunker a. , coulton w. , croft r. , di matteo t. , khandai n. , feng y. 2013 , arxiv:1302.1387 witt , a. n. , & gordon , karl d. 2000 , apj , 528 , 799 yajima , h. , umemura , m. , mori , m. , nakamoto , t. 2009 , mnras , 398 , 715 yue , b. , ferrara , a. , salvaterra , r. , chen , x. 2013 , mnras , in press ( arxiv1208.6234 ) zackrisson e. , bergvall , n. , leitet , e. 2008 , apj , 676 , l9 zackrisson e. , rydberg c .- e . , schaerer d. , stlin g. , tuli m. , 2011a , apj , 740 , 13 zackrisson e. , inoue a. k. , rydberg c .- e . ,
duval f. 2011b , mnras , 418 , l104 zackrisson e. et al .
2012 , mnras , 427 , 2212 zahn , o. , et al .
2012 , apj , 756 , 65 | the fraction of ionizing photons ( @xmath0 ) that escape from @xmath1 galaxies is an important parameter when assessing the role of these objects in the reionization of the universe , but the opacity of the intergalactic medium precludes a direct measurement of @xmath0 for individual galaxies at these epochs .
we argue , that since @xmath0 regulates the impact of nebular emission on the spectra of galaxies , it should nonetheless be possible to indirectly probe @xmath0 well into the reionization epoch . as a first step ,
we demonstrate that by combining measurements of the rest - frame uv slope @xmath2 with the equivalent width of the @xmath3 emission line , galaxies with very high lyman continuum escape fractions ( @xmath4 ) should be identifiable up to @xmath5 through spectroscopy with the upcoming james webb space telescope ( jwst ) . by targeting strongly lensed galaxies behind low - redshift galaxy clusters , jwst spectra of sufficiently good quality
can be obtained for @xmath6 galaxies at @xmath7 and for @xmath8 galaxies at @xmath5 .
dust - obscured star formation may complicate the analysis , but supporting observations with alma or the planned spica mission may provide useful constraints on the dust properties of these galaxies . |
the study of outflows in cataclysmic variables ( cvs ) was born in the iue era .
iue ultraviolet spectra of luminous cvs novalikes ( nls ) and dwarf novae ( dn ) in outburst showed high ionization uv transitions , particularly of civ @xmath01548,1552 , siiv @xmath01393,1402 and nv @xmath01238,1242 , appearing with very broad , blueshifted absorption components and p cygni profiles , shifting to pure emission in high orbital inclination systems @xcite .
examples of wind - dominated fuv spectra of cvs at different inclinations are given in figure 1 , which shows observations taken from the far ultraviolet spectroscopic explorer ( fuse ) data archive .
it was immediately noted that the line profiles were similar to those seen in early - type and luminous stars such as o stars and wolf - rayet stars , in which the lines were ( and are ) believed to originate in expanding winds driven by radiation pressure @xcite . as a result
, a wind origin for the uv lines of high accretion rate cvs was adopted early on .
iue spectroscopy revealed several properties of the cv winds .
the maximum blue edge velocities of the absorption troughs ( @xmath1 km s@xmath2 ) , which give a lower limit to the terminal velocity of the wind , were of order the escape velocity from the white dwarf ( wd ) accretor , suggesting that the outflows originate close to the wd @xcite .
the deepest absorption in the lines occured not at the blue edges of the lines , as is the case for early - type stars , but at lower velocities ( @xmath3 km s@xmath2 ) , suggesting that the wind is more slowly accelerated in cvs than in luminous stars @xcite . in eclipsing systems , the lines , unlike the continuum ,
were not occulted , indicating that the line - emitting region was extended relative to the bulk of the disk , with sizes of order the size of the occulting donor star @xcite .
a comparison of the strengths of the lines with model lines from spherically - symmetric , radiation pressure driven winds ( designed for luminous stars ) gave mass - loss rates in the wind from @xmath4
@xmath5 yr@xmath2 , or @xmath6 @xmath7 of the mass accretion rate in the disk @xcite . from the start
, models for cv winds have been based on the wind models for luminous stars : resonant line scattering from absorption of photons from the central source transfers momentum to ions ; as the wind expands outwards , the local resonance is redshifted , perpetuating the line scattering and the driving of the outflow .
( for a more detailed discussion of the physics of cv wind models , see proga , this volume . )
the first models assumed a radial outflow with a constant ionization structure emerging from the center of the disk and the wd @xcite .
the models were used to compare theoretical line profiles to the observed lines , in particular civ , the strongest wind feature in the uv spectrum .
the results revealed problems with the picture of a radial wind . in high inclination cvs
, the blueshifted absorption component of the wind lines disappears as the disk becomes more edge - on , but the radial models continue to show absorption at high inclination .
@xcite cited this as evidence for bipolarity , rather than spherical symmetry , in the outflow , although it was noted by @xcite that radial winds could still be present if the bipolarity was introduced by limb - darkening effects in the accretion disk .
mauche & raymond pointed out another significant problem with radial cv winds , however : radial winds `` see '' so much of the continuum from the wd and boundary layer ( bl ) that the wind is ionized beyond civ unless it is highly optically thick , which requires wind mass loss rates of @xmath81/3 of the mass accretion rate .
such a massive wind can not be radiatively driven , produces uv line fluxes an order of magnitude larger than observed values , and would absorb the soft x - ray continuum from the bl , which is inconsistent with observed x - ray fluxes .
on the basis of their models , the authors concluded that radial outflows were unlikely in cvs . as a result , the next generation of cv wind models adopted biconical , axisymmetric wind structures , with the winds being launched from the accretion disk .
@xcite developed a kinematical model of a rotating , biconical disk wind , taking into account radiation from the wd , the bl , and the accretion disk when calculating the local ionization structure of the ( constant temperature ) wind .
radiation transfer and line driving of the wind were treated using the standard wind modeling theory for luminous stars . by comparing theoretical line profiles from their models with those of radial winds
, they showed that biconical disk winds provide a better match to the profiles of civ wind lines at varying inclinations and do not suffer from the excessive absorption of bl photons and subsequent over - ionization of the wind as seen in radial wind models .
@xcite compared the biconical disk model line profiles to observed iue spectra of cvs .
they were able to match the civ profiles of the low inclination system , rw sex , and the eclipsing nl , rw tri , with wind mass - loss rates of order 1 15% of the mass accretion rate .
their models gave a typical scale for the civ line of 50
100 r@xmath9 in vertical height above the accretion disk . at this point ,
the theoretical and observational evidence both pointed to the presence of biconical accretion disk winds driven by radiation pressure in high accretion rate cvs .
much of the progress in understanding cv outflows continues to come from uv spectroscopic observations of luminous systems , although euv and x - ray observations of cvs have also provided information on the properties of the winds .
winds have also been cited as possible sources for features in optical lines , particularly non - keplerian emission in heii @xmath10 , but the evidence for wind signatures at optical wavelengths is generally indirect @xcite .
one exception is the optical spectrum of bz cam , which shows intermittent p cygni profiles in h@xmath11 and hei @xmath125876 @xcite .
the optical lines are complex blends of emission from the wind and the disk , however , which limits analysis of the wind behavior based on optical emission . in the uv
, the advent of sensitive , high time resolution uv spectrographs on hst and fuse has allowed direct tests of the predictions of disk wind models .
below , we examine several properties of cv winds that have resulted from observations in the past decade .
one observational signature of line - driven disk winds is that they rotate and that the winds preserve angular momentum as they leave the disk .
this is in contrast to hydromagnetically - driven winds , such as those seen in ysos , which maintain a constant angular velocity out to the alfvn radius and perhaps beyond . in at least two cvs ,
the observational evidence favors a rotating wind that conserves angular momentum .
@xcite compared the out of eclipse and mid - eclipse uv line profiles in iue observations of the nearly edge - on nl , v347 pup .
they showed that the civ wind line narrows in eclipse when the region of the wind near the disk surface that is responsible for the most of the rotational broadening of the line is occulted .
they demonstrated that the line profile variations during eclipse are consistent with the biconical disk wind model of @xcite when only orbital phase changes are invoked .
the civ emission line also narrows during eclipse in ux uma and a rotational signature can be seen in the eclipse light curves of different velocities in the line : the eclipse appears in the blue wing of the line before it appears in the red wing , and the eclipse is most shallow at the line center @xcite .
@xcite modeled the ux uma data high spectral resolution , time - resolved hst / ghrs observations of civ with a biconical disk wind model and were able to reproduce the pre- and mid - eclipse spectra and the velocity - resolved eclipse light curves with moderately collimated ( 65@xmath13 opening angle ) , rotating disk wind .
observations of eclipsing systems have also been instrumental in probing the vertical structure of the winds as they are launched from the disk .
the hst observations of the civ line in ux uma , in addition to exhibiting a rotational disturbance , show narrow , low - velocity absorption components that are occulted during eclipse .
the same narrow absorption components are observed in numerous other lines in the hst spectral range and also appear in phase - resolved fuse spectra of ux uma @xcite .
the eclipse behavior of the uv spectrum of ux uma is shown in figure 2 .
outside of eclipse , the spectrum is a complex blend of line emission , with no discernable continuum . at mid - eclipse ,
the spectrum is much cleaner , and is characterized by a flat continuum and strong , broad line emission from the unocculted portion of the wind .
the difference spectrum , which is the spectrum of the eclipsed light , shows that the eclipsed parts of the wind lines are narrow ( @xmath14 1000 km s@xmath2 ) absorption components centered on the rest wavelengths of the lines .
@xcite showed that to successfully model these absorption dips and their disappearance during eclipse requires the presence of a dense , low outflow velocity transition region between the bulk of the accretion disk and the fast wind .
the transition region , or disk chromosphere , may be part of the wind itself : the hydrodynamical disk wind models of @xcite show that when the line driving is dominated by radiation from the disk ( over that from the wd and bl ) , the wind has both a fast outflow component and a dense , low outflow velocity region near the disk surface .
recent observations also point to azimuthal structural variations in the disk winds .
time - resolved fuse spectra of rw sex show that the fuv wind lines are modulated on the orbital period @xcite . over the orbit , all of the wind lines in the fuv spectrum
ranging from lower - ionization transitions of ciii and niii to pv , svi , and ovi show the same behavior : a single - humped modulation in which the velocity of the deepest absorption of each line moves to the red , accompanied by an increase in the absorption depth , and a shift back .
the authors speculate on the source of the orbital phase modulation , which they link to an asymmetry in the structure of the disk .
orbital - phased modulations in wind lines have also been observed in the nls v795 her and v592 cas @xcite .
the source of the orbital changes in the wind lines remains unknown .
@xcite suggest that a disk tilt or warp could account for the modulation in rw sex , but this explanation is not favored for v592 cas , as the wind variations are modulated on the orbital period , but not on the known superhump period of the system .
orbital changes in the uv continuum and ( non - wind ) line spectra , particularly associated with the phases before eclipse , have also been seen in nls and dn in outburst @xcite .
the orbital effects are believed to be caused by interactions between the disk and the mass accretion stream from the donor star , resulting in a bulge in the outer disk or in stream overflow , similar to the source of x - ray dips in some x - ray binaries @xcite .
azimuthal variations in cv winds may be tied to the disk - stream interaction region , but the explicit link between outer disk and wind asymmetries has not yet been discovered . while some cv winds show orbital modulation , others are dominated by secular variability .
( oddly , there is little overlap between these two sets of systems . )
@xcite studied high time resolution hst / ghrs spectra of the nl , bz cam .
the spectra show strong and continuous variability in the absorption troughs , but not the emission components , of the p cygni lines of cii @xmath15 , ciii @xmath16 , si iii @xmath17 , siiv , nv , and civ . the lines vary together and can show optical depth changes of a factor of 5 over 1000
3000 km s@xmath2 of the absorption troughs on 100 sec time scales .
since all lines vary together , the variability is likely tied to density fluctuations in the wind rather than changes in the ionization state . in bz cam , then , the picture is that of a turbulent , stochastic wind , a contrast to the more stable winds from luminous stars .
the outflows in cvs are much more compact than those from stars prinja et al .
note that a wind travelling at 3000 km s@xmath2 can traverse the binary separation ( @xmath1850r@xmath9 ) in 2 3 min and may therefore reflect underlying disk fluctuations on short time scales .
this chaotic , unsteady outflow is also consistent with the behavior of the wind in the models of @xcite , although in other hydrodynamic models , a more steady outflow is seen @xcite .
some cvs do not show the short time scale variability exhibited by bz cam , but do show variable line profiles in the form of narrow absorption components to the lines . in the nls ix vel and v3885 sgr , very narrow ( @xmath19 km s@xmath2 ) ,
blueshifted ( @xmath20 km s@xmath2 ) absorption dips appear in the absorption troughs of several of the wind lines at times of strongest wind activity @xcite .
similar dips appear intermittently in the profiles of the ovi @xmath01032,1038 doublet lines of u gem in outburst @xcite . in u gem , the lines ( located at @xmath21 km s@xmath2 ) appear erratically and are not correlated with orbital phase ; sometimes , two dips appear in each of the doublet transitions .
the features are analogous to the discrete absorption components ( dacs ) seen in luminous star winds , although the cv absorption dips remain at constant velocity , while the dacs are observed to move slowly outward with time .
because the dips in the cv lines do not move , many of the models invoked to explain dacs will not apply for the cv dips .
a possible source for the dips is the formation of localized plateaus in the outflow velocity law , leading to pileup of ions . determining the formation mechanisms of winds in high accretion rate cvs
also depends on understanding cases in which the `` standard '' uv wind features do not appear .
one example is u gem in outburst : with the exception of the aforementioned absorption dips in ovi and perhaps a weak p cygni profile in the same , there is no sign of an outflow in the fuv spectra of u gem ; instead , the lines are generally narrow , low - velocity absorption features and sometimes are absent altogether @xcite .
one possibility is that the bl luminosity plays a role in the structure of the wind .
u gem is one of the few cvs that shows evidence for a luminous bl in outburst : euv spectra show a hot ( 140,000 k ) blackbody component whose luminosity , consistent with classical accretion disk theory , is comparable to that of the accretion disk @xcite .
the same euv observations show highly ionized resonant line transitions in emission .
the lines that appear are consistent with an origin in gas photoionized by the bl and are likely formed in a highly ionized wind .
further understanding of the ionization structure of winds will require comparisons of uv , euv , and x - ray spectra in different systems .
the lower inclination dn , ss cyg , has a low - luminosity bl , for example , and shows wind features in both x - ray and fuv spectra , but no euv wind lines ( mauche 2004 ; long et al .
2004 , in preparation ) .
good fits to the x - ray spectrum can be obtained with a wind model , but only if the individual ionizations fractions of the lines are allowed to vary arbitrarily . in the dn oy car in superoutburst , the euv spectrum shows strong wind emission in resonant transitions of several ions @xcite .
mauche & raymond note , however , that the radiation pressure required to drive the wind in oy car can not be met by current wind models , a point that is discussed further below . using time - tagged hst / stis observations of the nls ix vel and v3885 sgr , @xcite tested radiatively - driven wind models by comparing luminosity variations in the uv continuum and in the wind lines .
for a purely radiative wind , the strength of the wind lines should be directly correlated with the strength of the photoionizing continuum , but they found no such correlation in either system : while both the continuum and the lines varied with time , the wind had often weakened when the continuum was higher , and two epochs with the same continuum level in ix vel showed strong wind features and little to no wind emission at all , respectively .
@xcite concluded that the observations provided a serious challenge to the model of purely radiatively - driven disk winds .
the result was not entirely surprising , as there have been other indications that line - driving may not be the sole wind - formation mechanism at work in cvs .
one problem that continues to plague hydrodynamic wind models is an inability to match the observed strengths of the wind lines .
@xcite showed that when current determinations of mass accretion rates in luminous cvs are assumed , the hydrodynamic models underpredict the strength of the wind by an order of magnitude . as a result ,
additional mechanical wind drivers may be present in luminous cvs .
the possible role of centrifugal forces in driving disk winds was considered at an early date by @xcite , who also noted that hydromagnetically - driven outflows would strongly influence binary secular evolution as sinks of angular momentum .
in addition , @xcite showed that the presence of shocks in the wind can cause local density enhancements that lower the overall wind mass loss rate required to match observed wind line strengths .
work on incorporating non - radiative forces into cv wind models is only beginning ( see proga , this volume ) , but represent an important avenue for future studies of disk winds .
most of the modeling of cv winds has concentrated on one transition , civ @xmath01548,1552 . as a result
, our understanding of the thermal and ionization structure of disk winds remains limited .
the next steps in understanding the formation and behavior of cv disk winds will depend on simultaneously modeling the broad range of species and ionization levels that appear in x - ray and uv spectra of cv winds . to this end
, @xcite developed a monte carlo radiative transfer program that , given an input wd and accretion continuum spectrum , self - consistently calculates the thermal and ionization structure of the wind and generates synthetic spectra at every viewing inclination .
an example of the types of fit possible with their program is shown in figure 3 , which shows a disk and wind model fit to the outburst spectrum of z cam , obtained with the hopkins ultraviolet telescope .
a qualitatively good fit to the observed spectrum can be obtained with a model in which the disk has a mass accretion rate of @xmath22 m@xmath23 yr@xmath2 and the mass loss rate in the wind is @xmath24 m@xmath23 yr@xmath2 . in the case of relatively simple spectrum , such as the fuv spectrum of ss cyg in outburst , which is dominated by ovi ( see figure 1 ) , a good model fit can be achieved assuming a steady - state accretion disk and a disk wind carrying away @xmath25 of the accreted mass @xcite . the wind mass loss rate obtained from the model fit (
@xmath26 m@xmath23 yr@xmath2 ) is also within a factor of a few of the mass loss rate determined from chandra observations of ss cyg in outburst @xcite .
once the spectrum becomes more complex , however , with wind signatures appearing in many lines , as in the case of z cam above , fitting the entire spectrum and determining uniqueness of the fit parameters becomes challenging .
cv spectra are extremely complex and can contain emission from multiple components in the system , parameters for some of which remain poorly constrained .
nevertheless , the combination of high quality observations and sophisticated models offer a promising opportunity to quantify the structure and behavior of disk winds in cv systems .
outflows , in the form of moderately collimated winds , are closely tied to disk accretion in luminous cvs .
a combination of observations and modeling have shown that cv winds are bipolar , highly turbulent , and structurally complex .
radiative line - driving models for the source of the wind explain many of the observations , but additional , mechanical mechanisms may play a role in the structure and formation of the wind .
future observations and modeling will concentrate on determining the vertical and azimuthal distribution of the wind , its thermal and ionization structure , and the mechanisms that influence wind formation and behavior .
froning , c. s. , long , k. s. , & knigge , c. 2003 , apj , 584 , 433 froning , c. s. , long , k. s. , drew , j. e. , knigge , c. , proga , d. , & mattei , j. 2002 , in asp conf .
261 , the physics of cataclysmic variables and related outflows , eds .
b. t. gnsicke , k. beuermann , & k. reinsch ( san francisso : asp ) , 337 hartley , l. e. , drew , j. e. , long , k. s. , knigge , c. , & proga , d. 2002 , mnras , 332 , 127 hassall , b. j. m. , pringle , j. e. , schwarzenberg - czerny , a. , wade , r. a. , whelan , j. a. j. , & hill , p. w. 1983 , , 203 , 865 heap , s. r. , boggess , a. , holm , a. , klinglesmith , d. a. , sparks , w. , west , d. , wu , c. c. , boksenberg , a. , willis , a. , & wilson , r. 1978 , nature , 275 , 385 holm , a. v. , panek , r. j , & schiffer , f. h. , iii 1982 , apj , 252 , l35 honeycutt , r. k. , schlegel , e. m. , & kaitchuck , r. h. 1986 , , 302 , 388 long , k. s. , froning , c. s. , g " ansicke , b. , knigge , c. , sion , e. m. , & szkody , p.2003 , , 591 , 1172 long , k. s. & knigge , c. 2002 , apj , 579 , 725 long , k. s. , mauche , c. w. , raymond , j. c. , szkody , p. , & mattei , j. 1996 , apj , 469 , 841 marsh , t. r. & horne , k. 1990 , , 349 , 593 mason , k. o. , drew , j. e. , cordova , f. a. , horne , k. , hilditch , r. , knigge , c. , lanz , t. , & meylan , t. 1995 , , 274 , 271 mauche , c. w. 2004 , apj , 610 , 422 mauche , c. w. 1991 , apj , 373 , 624 mauche , c. w. & raymond , j. c. 2000 , , 541 , 924 prinja , r. k. , knigge , c. , witherick , d. k. , long , k. s. , & brammer , g. 2004 , mnras , in press prinja , r. k. , long , k. s. , froning , c. s. , knigge , c. , witherick , d. k. , clark j. s. , ringwald , f. a. 2003 , mnras , 340 , 551 prinja , r. k. , ringwald , f. a. , wade , r. a. , & knigge , c. 2000 , mnras , 312 , 316 proga , d. , stone , j. m. , & drew , j. e. 1998 , mnras , 295 , 595 | outflows in the form of fast winds from the accretion disk appear in nearly all high accretion rate cataclysmic variables ( cvs ) , both novalikes ( nls ) and dwarf novae in outburst ( dn ) .
the primary signatures of cv winds are broad , blueshifted absorption features and p cygni profiles that appear in the ionized metal lines of lower inclination systems and the broad emission lines that appear in eclipsing systems .
much progress has been made in our understanding of the structure and behavior of cv outflows based on data obtained with x - ray , euv , and uv telescopes and the kinematical models developed to fit the observations .
our current picture of cv outflows is that they are bipolar , rotating winds driven off the accretion disk primarily by radiative line driving from disk , boundary layer , and white dwarf radiation .
other mechanical forces may play a role in the formation of cv winds .
the winds are highly turbulent and show complex vertical and azimuthal structure .
future work will focus on determining the physical , thermal and ionization structure of the winds and probing the mechanisms that govern wind formation and behavior . |
in a recent communication m.m . block and f. halzen @xcite ( hereafter referred to as bh ) have presented some critical comments on our analysis @xcite ( hereafter fms ) .
some points raised by these authors have already been addressed and discussed in our subsequent work @xcite ( hereafter fms2 ) , available as arxiv since august 16 , 2012 . in this note
we reply to the bh criticisms directed to fms , recalling also some aspects of interest presented in fms2 . first , to facilitate the discussion and the reference to each part of the bh commentaries , some explanations on the organization of this reply are in order . excluding the acknowledgments and references , bh arxiv comprises four pages and the effective criticisms to fms
correspond to approximately one page .
all the rest of the manuscript ( three pages ) largely overlap with their previous work @xcite ( as advised in the arxiv admin note " @xcite ) .
we shall focus on this 25 % material , in our section [ s2 ] .
although not characterized as criticisms , the rest of the bh reproduces their former work on the subject , as a kind of lesson to be learned . in this respect ,
a discussion on some aspects of the fms analysis and a brief commentary on the bh former work are presented in section [ s3 ] .
our conclusions and final remarks are the contents of section [ s4 ] .
the content of the criticisms to fms , presented in bh , can be divided in three blocks , one block referring to the @xmath0 information ( page 1 in bh ) , another block referring to statistical probabilities ( page 2 ) and the last one to predictions at 7 tev ( page 2 ) . in
what follows , each block will be treated as a subsection , in which we first situate and summarize the commentary , or quote it explicitly , and then present our reply .
_ - commentary _
the first effective criticism appears in page 1 , fourth paragraph of the section _ introduction_.
it concerns the fact that in fms the @xmath0 information was not used in global fits with the total cross section data . according to them , a statement alluded to ( but _ not _ carried out ) in appendix ... " .
they also add : ... in their appendix , they give a rather cumbersome evaluation using their variant 3 model , to _ separately evaluate _ @xmath0 .... " _ - our reply _ in fms , the analysis has been based only on the @xmath2 data ( without the inclusion of the @xmath0 information ) for the six reasons explained there , which we consider as six facts .
however , addressing the comments by one of the _ three referees _ in the submission to the braz .
, we have included appendix a in a revised version . in this appendix
we have shown that , even in the case of the largest values of the exponent @xmath3 ( method 1 , v3 and method 2 , v5 ) , the _ predictions _ for @xmath4 are in agreement with the experimental information . to connect @xmath5 and @xmath4 in an analytical way
, we have used singly - subtracted derivative dispersion relations in the operational form introduced by kang and nicolescu @xcite ( also discussed in @xcite ) .
in particular we have obtained an _ extended _ analytical result for the case of @xmath3 as a _ real _ parameter ( equations a7 and a8 in appendix a of fms ) . in respect to the effect of the @xmath0 information in data reductions , we have stated at the end of appendix a ( the references that follows concern fms ) :
finally , we recall that in simultaneous fit to @xmath2 and @xmath0 the subtraction constant affects both the low- and high - energy regions [ 47,48 ] .
that is a consequence of the strong correlation among the subtraction constant and all the other physical free fit parameters .
we plan to discuss this consequence and other aspects of the fit procedures in a forthcoming paper . "
also , in the last paragraph of our conclusions ( third sentence ) we added : these are aspects that we expect to consider in a future work , since they may provide information that is complementary to the results here presented . " in fact , in the fms2 we have extended the fms analysis in several aspects as referred to and outlined in the introduction of fms2 . in special , not only individual but also novel simultaneous fits to total cross section and @xmath0 information have been developed , leading to solutions with @xmath3 greater than 2 , _ despite the constraint involved_. the origin and role of the subtraction constant have also been discussed in some detail .
therefore , we see no sense in depreciating statements like alluded but not carried out " or they give a rather cumbersome evaluation " .
the second criticism in bh appears in page 2 , section _ statistical probabilities _ and involves five paragraphs , four with criticisms ( left column ) and the final one with the conclusions ( right column ) .
the main point here concerns the use of the integrated probability @xmath6 to punctually analyze the reliability of the fit results .
we shall treat each paragraph separately and in sequence .
however , before that , some aspects on our fit procedures and on statistical analysis demand a few comments for further reference .
@xmath7 introductory remarks - _ on the fit procedures _ the parametrization for the total cross section used in fms , introduced by amaldi _
@xcite , reads @xmath8^{-b_1 } + \ , \tau \ , a_2\ , \left[\frac{s}{s_l}\right]^{-b_2 } + \alpha\ , + \beta\ , \ln^{\gamma } \left(\frac{s}{s_h}\right),\end{aligned}\ ] ] [ eq1 ] where @xmath9 = -1 ( + 1 ) for @xmath1 ( @xmath10 ) scattering and @xmath11 = 1 gev@xmath12 is fixed .
the dependence is linear in four parameters ( @xmath13 , @xmath14 , @xmath15 and @xmath16 ) and nonlinear in the other four ( @xmath17 , @xmath18 , @xmath3 and @xmath19 ) .
as stated by bevington and robinson @xcite ( section 8.2 searching parameter space ) and also quoted recently in @xcite , _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ fitting nonlinear functions to data samples sometimes seems to be more an art than a science . " _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ nonlinear data reductions are not a trivial task .
they demand a methodology for the choice of the initial ( feedback ) values of the free parameters .
our strategy has been to test a _ grid _ of different ( physical ) feedback values for the free fit parameters so as to check the stability of the results , as shortly recalled in what follows . in both cases , fms and fms2 ,
we have considered as feedback the results already found by the particle data group ( pdg ) , which uses the standard compete collaboration highest - rank parametrization ( @xmath3 = 2 , a fixed parameter ) @xcite . in fms
we used the 2010 pdg edition @xcite and in fms2 the updated result from the 2012 pdg edition @xcite .
although applied to only a _ subset _ of the dataset analyzed in pdg , we understand that with these conservative choices " ( @xmath3 = 2 ) , we start with reasonable stable solutions ( already found by the pdg ) . with this input
we are then able to investigate possible _ departures from this solution _ in the case of @xmath3 as a free parameter ( including or not the totem datum @xcite , as done in fms and including this point in all fits , as done in fms2 ) . _ in addition _ , in order to investigate the effect of the feedback values in the fit results , we have considered another distinct choices in both fms and fms2 and 3 different versions in each case of fms ( referred to as six variants and a direct fit ) . as explicitly quoted in fms and fms2 , our data reductions have been carried out with the objects of the class tminuit of the root framework @xcite .
the statistical interpretations of the fit results , as well as aspects related to the error matrix , correlation matrix and analytical error propagation , have been based on the bevington - robinson book @xcite . for further reference , we recall that the error ( or covariance ) matrix provides the variances ( diagonal ) and covariances ( out of diagonal ) associated with each free fit parameter . the symmetric correlation matrix gives a measure of the correlation between each pair of free parameter through a coefficient with numerical limits @xmath20 1 ( full correlation ) and 0 ( no correlation ) . in the minimization program
@xcite a confidence level of one standard deviation was adopted in all fits ( up = 1 ) . in each test of fit , successive
running of the migrad have been considered ( up to 200 calls in fms and up to 2,000 calls in fms2 ) , until full convergence has been reached , with the smallest fcn ( @xmath21 ) and edm ( estimated distance to minimum ) , specifically edm @xmath22 ( adequate for the one sigma cl ) .
in addition , the error in the parameters should not exceed the central value . among _ several different tests _
, we have _ selected _ , under the above criteria , the seven variants presented in fms . these variants are related to two different choices for the input values for all fits , denoted method 1 and method 2 in fms . for further reference and clarity , we summarize and situate below the structure of the grid considered in fms ( v stands for variant and df for direct fit , also a variant ) . method 1 .
initial feedback values from pdg 2010 ( table 1 , second column ) : @xmath23 method 2 .
initial feedback values distinct from method 1 ( table 3 , second column ) : @xmath24 in the six variants ( including the df ) we have investigated the effects of fixing or not the three fundamental parameters directly related with the energy dependence , namely @xmath17 , @xmath18 and @xmath3 . for the reggeon intercepts we have tested either _ ad hoc _ fixed values 1/2 or fixed ( central ) values from spectroscopic data ( chew - frautschi plots ) .
we stress that these results constitute final solutions , selected under the above mentioned criteria .
therefore our strategy in fms ( and fms2 ) did not involve extremely detailed use of different routines to possibly reach an absolute minimum .
our point has been to investigate a _ grid _ of reasonable physical choices for feedbacks and variants for two ensembles , obtaining solutions through standard running of the minuit . despite of these possible limitations
, we have found several reasonable consistent solutions with @xmath3 greater than 2 and that has been the only essential point raised in fms ( and developed also in fms2 )
. - _ on statistics _ in fms and fms2 ,
following the pdg procedure , the dataset include statistical and _ systematic errors _ , added in quadrature . in our opinion ,
the inclusion of systematic errors puts certain limits in a _ full _ statistical interpretation of the fit results .
in fact , the @xmath21 test for goodness of fit is based on the assumption of a _
gaussian error distribution _ @xcite .
although statistical uncertainties are considered to follow this distribution , that , certainly , is not the case for systematic uncertainties , which are equally probable quantities .
therefore , we understand that a _ full _ statistical interpretation of data reductions including systematic uncertainties has a somewhat limited validity , specially in what concerns integrated probability ( due to the inclusion of equally probable quantities ) . in the fms and fms2 , the corresponding dof ( @xmath25 ) and @xmath26 for each fit has been displayed only to shown that they constitute reasonable ( acceptable ) statistical results .
the condition of reduced @xmath21 closest to 1.0 has been one of the criteria used to select a given result , but not the only one .
attempts to a full statistical interpretation of the results , mainly in terms of integrated probabilities , may lead to questionable conclusions , as discussed in what follows .
at last , it is important to note that the focus in fms does not concern comparison among models ( or variants ) , but between two ensembles , with or without the totem datum . also ,
when we refer to _ statistically consistent results _ for @xmath3 = 2 or @xmath3 above 2 , we mean that _ the corresponding numerical result for @xmath3 is consistent within their uncertainty_. let us now treat each paragraph from section _ statistical probabilities _ in bh .
we shall adopt here their notation ( lower index 1 for the @xmath27 = 1.8 tev ensemble and lower index 2 for the @xmath27 = 7 tev ensemble ) .
@xmath7 first paragraph - commentary bh discuss our results in table 1 of fms ( @xmath27 = 1.8 tev ) , comparing the data reductions that can be summarized as follows : df@xmath28 : @xmath29 @xmath25 = 156 , @xmath29 @xmath26 = 0.931 , @xmath29 @xmath30 = 0.721,@xmath29 case of @xmath3 = 2 ( fixed ) @xmath31v1 : @xmath29 @xmath25 = 155 , @xmath29 @xmath26 = 0.937 , @xmath29 @xmath32 = 0.701 , @xmath29 case of @xmath3 free according to block and halzen : ... we get the somewhat strange result that fms have a _ ... better , somewhat more reliable _
fit when they fix the value of @xmath3 at 2 , ... , than they allow it to float , suggesting perhaps that the _ true _
minimum @xmath21 was not achieved in their minimization process . "
they state in the last sentence : in any event , fms concluded that the value @xmath3 = 2 was correct for the energy interval 5 @xmath33 1,800 gev . " - our reply : the v1 result has been obtained from df@xmath28 as feedback .
based on bevington and robinson @xcite ( chapter 11 ) and in our above comments on statistics , for @xmath34 155/156 we do not think that @xmath35 0.72 @xmath36 @xmath37 0.70 strictly implies in a better fit , namely that df@xmath28 could be more reliable " than v1 .
in fact , let us compare the values of the free parameters in both fits in table 1 of fms ( third and fourth columns ) .
note that the central values of the parameters are identical up to 3 figures , except for @xmath16 ( 0.264 and 0.263 ) and @xmath19 ( 12.0 and 12.2 ) .
all the parameters in both fits are consistent within their errors , leading to the conclusion that both results are , effectively , equivalent ( the corresponding curves in figures 1 and 2 of fms overlap ) .
we understand that in this nonlinear fit , based on the same dataset ( @xmath27 = 1.8 tev ) , to let free _ only one parameter _ ( @xmath3 ) does not allow the punctual statistical interpretation by block and halzen ( we shall return to this point in the discussion on the fourth paragraph ) .
the very small differences in the central values of the parameters , and those in the errors , are associated with the _ correlations among all the fit parameters _ , resulting , in this particular case , in a small decrease of the probability when @xmath3 is let free .
the correlation coefficients associated with both fits are displayed in table [ t1 ] : df@xmath28 above the diagonal of the table and v1 below the diagonal ( we shall return to this point in what follows ) .
perhaps there can be some small differences in reaching a true minimum , but , in our opinion , that does not invalidate our results and interpretation .
.correlation coefficients from the ( symmetric ) correlation matrices @xcite associated with df@xmath28 and v1@xmath28 results .
the off - diagonal coefficients from df@xmath28 are displayed above the diagonal of the table ( not filled ) and those from v@xmath28 , below that diagonal . [ cols="^,^,^,^,^,^,^,^,^,^ " , ] [ t3 ] from figure 1 , the _ predictions _ by the compete 2002 and pdg 2010 are in agreement with the totem result , but not the data reduction from the pdg 2012 , which includes this point in the updated dataset : the curve and uncertainty region lie below the totem lower uncertainty bar , _ corroborating _ , therefore , _ the results and conclusions previously presented in fms_. note also that these results ( figure 1 ) corroborate another conclusion in fms , relating the totem point with the highest cosmic - ray estimates for the @xmath1 total cross section ( fly s eye collaboration @xcite and pierre auger collaboration @xcite ) .
in fact , in what concerns our fit results with the amaldi _
_ parametrization ( and in the particular case of figure 1 with @xmath3 = 2 , namely the compete highest - rank result ) , there is no agreement among these three points : curves in consistency with the totem datum lie above the central values of the cosmic - ray estimations and the same is true in the inverse sense .
based on all these facts , we understand that , once included in the dataset , it is , at least , not obvious the 7 tev totem result can be described by a standard @xmath38 $ ] leading dependence and that was the essential discovery " in fms . from our analysis ,
this effect is related with the fundamental correlation between @xmath3 and the scale factor @xmath19 ( and also with the subtraction constant @xmath39 in case of simultaneous fits to @xmath2 and @xmath0 data : see appendix a in fms2 on the correlation matrices ) .
note , from table [ t3 ] , that in the compete case , @xmath40 34 gev@xmath12 and in the pdg 2012 edition , @xmath40 16 gev@xmath12 ( below , therefore , @xmath41 25 gev@xmath12 ) : the effect of these differences can be seen in fig .
1 . in particular , _ once included in the dataset _ , it is not expected the 7 tev totem datum might be described for fixed @xmath3 = 2 and @xmath42 , as is the case in bh .
the totem collaboration has already obtained three new high - precision measurements of the total cross section at 7 tev , through different methods and techniques @xcite .
all the measurements are consistent within their uncertainties and therefore confirm the first result they have obtained ( which has been used in fms and fms2 ) . in this respect , we can advance that the inclusion of these three points in our dataset leads also to solutions with @xmath3 greater than two @xcite , corroborating the conclusions in fms .
however , the asymptotic ratio between elastic / total - cross - section , discussed in fms2 , is still under investigation .
a luminosity - independent measurement at 8 tev has been also reported , indicating @xmath43 101.6 @xmath20 2.9 mb @xcite .
this value lies above the prediction in fms2 , namely 98.7 @xmath20 1.0 mb , for @xmath3 = 2.346 @xmath20 0.013 and @xmath19 = 0.383 @xmath20 0.041 gev@xmath12 @xcite . at this point
, we could conjecture ( if not speculating ) on the implication of a possible increase of @xmath2 faster than @xmath44 .
in contrast with an effective violation of the froissart - martin bound , a fast rise of the total cross section might also be associated with some local effect at the lhc energy region , so that , asymptotically , the bound remains valid .
a faster - than - squared - logarithm rise points also to the possibility of a power - like behavior @xcite , which has always been and important and representative approach ( see , for example , the unitarized model @xcite ) .
these conjectures are not in disagreement with the recent theoretical arguments by azimov @xcite . here
we present some critical comments on the bh parametrization , the aspen model and the bh results for the total , elastic and inelastic cross sections .
the bh analytical parametrization for the total cross sections , used in global fits to @xmath2 and @xmath0-values from @xmath1 ( + ) and @xmath10 ( - ) scattering , reads @xcite @xmath45^{\mu - 1 } \ , \pm \ , \delta \ , \left[\frac{\nu}{m}\right]^{\alpha - 1 } + c_0\ , + c_1\ , \ln\left(\frac{\nu}{m}\right)\ , + c_2\ , \ln^2 \left(\frac{\nu}{m}\right),\end{aligned}\ ] ] [ eq2 ] where @xmath25 is the laboratory energy , @xmath46 is the proton mass and , in terms of the c.m .
energy , @xmath47 .
comparison with the parametrization used in fms , eq .
( 1 ) , shows that , for @xmath3 = 2 ( fixed ) , both forms have the same _ analytical structure_. however , the striking difference between fms and bh approaches concerns the number of fixed and free parameters and mainly , in the bh case , the way the parameters are fixed , as discussed in what follows . in the fms analysis , except for a dimensional notation choice , namely @xmath11 = 1 gev@xmath12 , all the 8 parameters involved have been treated as free or fixed in the 6 variants considered .
that is , there is no _ ad hoc _ fixed parameters , except for particular variant tests . in all cases ,
the uncertainties in the free parameters have been explicitly given .
moreover , by letting free different parameters in a _ nonlinear _ fit , we are able to investigate all the correlations involved , the variances , covariances and , as a consequence , the global uncertainties in all fitted and predicted quantities , as done in fms2 . in the bh approach , besides the arbitrary fixed energy scale ( corresponding to @xmath48 at both low and high energy regions ) , among the 7 parameters in eq .
( 2 ) , 5 are fixed ( without uncertainties ) and only 2 are free in data reductions .
specifically , from table iii in @xcite : - fixed parameters : @xmath49 = 31.10 mb , @xmath50 = 0.5 , @xmath51 = - 28.56 mb , @xmath15 = 0.415 and @xmath52 = 37.32 mb ; - fitted parameters : @xmath53 = - 1.440 @xmath20 0.070 mb and @xmath54 = 0.2817 @xmath20 0.0064 mb . as a consequence ,
the parametrization is linear in any reduction to @xmath2 data , leading to unique solution @xcite .
that , obviously , contrasts with eq .
( 1 ) and with the strategies in fms and fms2 , as commented above . in bh the reggeons intercepts are fixed , corresponding to @xmath17 = 0.5 and @xmath18 = 0.585 in eq .
( 1 ) , which is not in agreement with the spectroscopic data ( chew - frautschi plots ) and scattering fit results , as obtained by several authors @xcite .
since this assumption permeated the intermediate and low energy region , in our opinion , it puts some limits on the reliability of a formal connection with the finite energy sum rules at low energy .
the fixed parameters do not allow the study of correlations and their effects in the fitted and , most importantly , in the predicted quantities .
we shall return to this point in the next subsection on the aspen model .
we also note that , although ( 1 ) and ( 2 ) have the same analytical structure , the bh and fms high energy formulations for @xmath3 = 2 are not equivalent , even in the cases ( variants ) with @xmath17 and @xmath18 fixed .
in fact , in fms all the other parameters are free and the fit extends _ simultaneously _ to both low- and high - energy data ( @xmath55 and @xmath56 contributions in fms ) . as a consequence there
is strong correlations among the parameters from both @xmath55 and @xmath56 ( see , for example , the v4@xmath28 coefficients in table ii in the case of @xmath13 and @xmath19 , @xmath15 and @xmath16 ) .
since that is not the case in bh formulation ( @xmath49 and @xmath51 , corresponding to @xmath13 and @xmath14 , are fixed ) , we see no correspondence between the high energy formulations , as stated in bh ( after eq .
( 7 ) in that paper ) .
in the bh approach the main hypothesis concerns the imposing that the fits to the high energy data smoothly join the cross section and energy dependence obtained by averaging the resonances at low energy " @xcite . in the fms analysis , on the contrary ,
we have tried to identify possible high - energy effects that may be unrelated to the trends of the lower - energy data ... " @xcite .
therefore , the assumptions , approaches and strategies in fms and bh are completely different .
we see no reason for the comparative discussion presented in bh . in the aspen model @xcite two fundamental quantities ,
the mass scale @xmath57 and the coupling constant @xmath58 , are unknown parameters , fixed to _ ad hoc _ values of 600 mev and 0.5 , respectively in order to obtain best fits in data analysis .
the @xmath59 parameter , from the gluon structure function , is also fixed at 0.05 .
the two fundamental parameters , @xmath57 and @xmath58 , have been reinterpreted by luna _ et al .
_ in the context of a dynamical gluon mass ( dgm ) approach @xcite . that has allowed connections with nonperturbative qcd , as expected in the soft sector represented by the elastic scattering processes . in the dgm approach ,
two essential parameters are the dynamical gluon mass scale @xmath60 and the soft pomeron intercept @xmath59 . more recently and most important for our purposes , fagundes ,
luna , menon and natale ( flmn ) have developed a detailed analysis on the influence in the evaluated quantities associated with physical intervals for the @xmath60 and @xmath59 parameters @xcite .
moreover , in a similar way as done recently by achilli _
@xcite , flmn have established bounds and uncertainty regions in all evaluated quantities ( fitted and predicted ) , in accordance with the relevant physical intervals for @xmath60 and @xmath59 .
the main conclusion is that the uncertainty regions play a crucial role in the energy region above that used in the data reductions . in other words ,
the relevant intervals for each parameters affect substantially the high - energy predictions and they can not be fixed at _ ad hoc _ values , without a clear physical justification .
the above conclusion , expressed in @xcite , puts serious limits on the predictions of phenomenological models constructed on the basis of fixed parameters , whose numerical values do not have an explicit justification and/or whose _ consequences in the evaluated quantities are not investigated or even discussed_. in our opinion , these limitations are present in the foundations of both the bh analytical parametrization and the aspen model . in order to connect their global analytical ( empirical )
fits to @xmath2 and @xmath0 data with the inelastic cross section , @xmath61 , block and halzen use the predictions from the aspen model , in a kind of hybrid approach ( semi - empirical or perhaps semi - phenomenological ) @xcite .
the model prediction is parametrized by an analytic expression ( eq .
( 8) in @xcite ) , with fixed mass and power parameters , without any reference to the uncertainties in the fit parameters . with this hybrid approach , from the model evaluation of @xmath61 and the fit result for @xmath2
they infer that @xmath62 a result statistically consistent with the black - disc limit .
as commented in fms2 , we understand that the least ambiguous way to estimate the inelastic cross section is through the s - channel unitarity , @xmath61 = @xmath2 - @xmath63 , as has been done by the totem collaboration @xcite . that avoids the model dependence involved in direct estimation of @xmath61 , due to single and double diffraction contributions . in this respect
, it may be interesting to note the prediction in bh for the inelastic cross secion at 7 tev ( fig . 3 in @xcite ) : the uncertainty region barely reaches the lower error bar in the totem result .
moreover , for the total cross section , the central value of the prediction reaches the low extreme bar of the totem result ( and the uncertainty region , above the central value , lies through the lower error bar of the totem point ) .
it should be also noted that as a one - channel eikonal formalism , the aspen model does not take explicit account of the inelastic diffractive contributions . in what concerns the bh total cross section prediction at 57 tev and the recent estimation of this quantity by the pierre auger collaboration @xcite ,
we stress a peculiar statement in @xcite ( also present in @xcite ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ in particular , the agreement with the new highest energy ( 57 tev ) experimental measurement of both @xmath2 and @xmath61 is striking . " _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ first , certainly these are not _ experimental measurements _ , due to the strong model dependence involved , as already discussed by several authors and also in @xcite .
they constitute estimations of these quantities because they are based on extrapolations from models tested only in lower energies . moreover ,
recall that the estimation for the @xmath1 total cross section at 57 tev reads @xcite @xmath64\ , \mathrm{mb } $ ] . as discussed in fms2
( and also in our section ii.b ) , systematic and , here , theoretical ( glauber ) uncertainties are equally probable quantities ( do not follow a gaussian distribution ) .
that means the above central value is equally likely to lie in any place limited by the corresponding nonstatistical uncertainties , namely around @xmath20 23 mb .
therefore , we see no physical meaning in a statement referring to a striking agreement " . at last
, we understand that the bh analysis and the aspen model represent important contributions in the investigation of the high - energy elastic hadron scattering
. however , given the model character and _ ad hoc _ assumptions involved we see no conclusive evidence that the bh results constitute an unique and exclusive solution for the rise of the total cross section at high energies . in this respect , a reanalysis by block and halzen , including in their dataset the four high - precision totem measurements and presenting their linear fit prediction at higher energies , might be instructive .
in fms we have presented a study on the rise of the total hadronic cross section , with focus on the recent 7 tev totem result .
the analysis was based on a specific class of analytical parameterization , eq .
since the effects of all parameters involved have been considered , we were faced with a nonlinear data reduction , which constitutes a non - trivial problem .
our strategy in fms ( and fms2 ) has been to investigate a grid of different physical choices for feedbacks and the corresponding solutions . in both fms and fms2 , beyond a second distinct possibility , we have considered conservative " choices for the initial values , namely results previously obtained by the pdg with @xmath3 = 2 .
the data reductions have been developed through standard running of the minuit , namely the default minuit error analysis .
perhaps , in the fits presented in fms a true minimum had not been reached in some cases .
we understand that this does not invalidate the _ general and global conclusions _ from our grid strategy " approach ( summarized in section ii and , the corresponding results , in fig . 8 of fms ) .
anyway , further analysis , looking for optimizations in the use of the minuit code , including also tests with other computational tools for nonlinear fits ( as the subroutine minos @xcite ) , are certainly important and we intend to implement that .
the lack of unique solutions in our nonlinear fit procedures has been referred to in fms and fms2 .
we have never claimed to have obtained _ unique or absolute solutions_. in particular , we have concluded in fms ( the equations refer to that paper ) , _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ from our data reductions through parametrization ( 3 - 5 ) to @xmath1 and @xmath10 scattering above 5 gev , including the 7 tev totem result , we conclude that the total hadronic cross section may rise faster than @xmath44 at high energies . " _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ our results suggest that the energy dependence of the hadronic total cross section at high energies still constitutes an open problem . " _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ and in fms2 we have stressed , _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we also emphasize that our results represent possible consistent statistical solutions for the behavior of the total cross section , but do not correspond to unique solutions . " _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ it is important to note that at the highest energies , once treated as free fit parameters , the exponent @xmath3 and the scale factor @xmath19 are correlated in nonlinear data reductions . in the lack of formal and/or theoretical information on the value of this scale factor and without the _ canonical assumption _
@xmath3 = 2 , different solutions , in agreement with the experimental data , can be obtained for the leading @xmath65 contribution .
therefore , we understand that the analytical description of the rise of the total hadronic cross section at high energies still constitutes an open problem , demanding further investigations .
new high - precision data to be published and to be obtained by the totem collaboration at 8 - 14 tev are expected to shed light on the subject .
at last , based on both the @xmath66 " by block and halzen and this @xmath67 " , we see no evidence that invalidate the analysis , results , conclusions and the honest statements expressed in fms .
we are thankful to j.r.t .
de mello neto for useful discussions .
research supported by fapesp ( contracts nos . 11/15016 - 4 , 11/00505 - 0 , 09/50180 - 0 ) .
azimov , froissart bounds for amplitudes and cross sections at high energies , proc .
xlv winter school of pnpi , edited by v.t .
kim ( pnpi , st .
petersburg , russia , 2011 ) pp .
20 - 26 , arxiv : 1204.0984 [ hep - ph ] .
fagundes , m.j . menon and p.v.r.g .
silva , forward amplitude analysis and the recent 7 tev totem measurements , presented at xxiv reunio de trabalho sobre interaes hadrnicas , centro brasileiro de pesquisas fsicas ( cbpf ) , 3 - 4 december 2012 , rio de janeiro , brazil ( slides available at http://www.ifi.unicamp.br/@xmath69menon/24-rtih ) . | a reply to the above mentioned commentary by m.m . block and f. halzen on our quoted work is presented .
we answer to each point raised by these authors and argument that our data reductions , strategies and methodology are adequate to the nonlinear - fit - problem in focus . in order to exemplify some arguments , additional information from our subsequent analysis
is referred to . a brief commentary on the recent results by block and halzen is also presented .
we understand that this reply gives support to the results and conclusions presented in our quoted work .
pacs : 13.85.-t hadron - induced high- and super - high - energy interactions , 13.85.lg total cross sections , 11.10.jj asymptotic problems and properties * table of contents * i. introduction \ii .
reply to the criticisms ii.a on the @xmath0 information ii.b statistical probabilities ii.c prediction of the 7 tev total @xmath1 cross section \iii .
discussion iii.a on the fms analysis and results iii.b brief commentary on bh analysis and results \iv .
conclusions and final remarks |
since the seminal work of koenker and bassett ( 1978 ) , quantile regression has received substantial scholarly attention as an important alternative to conventional mean regression . indeed , there now exists a large literature on the theory of quantile regression ( see , for example , koenker ( 2005 ) , yu _ et al_. ( 2003 ) , and buchinsky ( 1998 ) for an overview ) .
notably , quantile regression can be used to analyse the relationship between the conditional quantiles of the response distribution and a set of regressors , while conventional mean regression only examines the relationship between the conditional mean of the response distribution and the regressors .
quantile regression can thus be used to analyse data that include censored responses .
powell ( 1984 ; 1986 ) proposed a tobit quantile regression ( tqr ) model utilising the equivariance of quantiles under monotone transformations .
hahn ( 1995 ) , buchinsky and hahn ( 1998 ) , bilias _ et al_. ( 2000 ) , chernozhukov and hong ( 2002 ) , and tang _ et al_. ( 2012 ) considered alternative approaches to estimate tqr .
more recent works in the area of censored quantile regression include wang and wang ( 2009 ) for random censoring using locally weighted censored quantile regression , wang and fygenson ( 2009 ) for longitudinal data , chen ( 2010 ) and lin _ et al_. ( 2012 ) for doubly censored data using the maximum score estimator and weighted quantile regression , respectively , and xie _ et al_. ( 2015 ) for varying coefficient models .
in the bayesian framework , yu and stander ( 2007 ) considered tqr by extending the bayesian quantile regression model of yu and moyeed ( 2001 ) and proposed an estimation method based on markov chain monte carlo ( mcmc ) .
a more efficient gibbs sampler for the tqr model was then proposed by kozumi and kobayashi ( 2011 ) .
further extensions of bayesian tqr have also been considered .
kottas and krnjaji ( 2009 ) and taddy and kottas ( 2012 ) examined semiparametric and nonparametric models using dirichlet process mixture models .
reich and smith ( 2013 ) considered a semiparametric censored quantile regression model where the quantile process is represented by a linear combination of basis functions . to accommodate nonlinearity in data , zhao and
lian ( 2015 ) proposed a single - index model for bayesian tqr .
furthermore , kobayashi and kozumi ( 2012 ) proposed a model for censored dynamic panel data . for variable selection in bayesian tqr ,
ji _ et al_. ( 2012 ) applied the stochastic search , alhamzawi and yu ( 2014 ) considered a @xmath2-prior distribution with a ridge parameter that depends on the quantile level , and alhamzawi ( 2014 ) employed the elastic net .
as in the case of ordinary least squares , standard quantile regression estimators are biased when one or more regressors are correlated with the error term .
many authors have analysed quantile regression for uncensored response variables with endogenous regressors , such as amemiya ( 1982 ) , powell ( 1983 ) , abadie _ et al_. ( 2002 ) , kim and muller ( 2004 ) , ma and koenker ( 2006 ) , chernozhukov and hansen ( 2005 ; 2006 ; 2008 ) , and lee ( 2007 ) . extending the quantile regression model to simultaneously account for censored response variables and endogenous variables
is a challenging issue . in the case of the conventional tobit model with endogenous regressors ,
a number of studies were published in the 1970s and 1980s , such as nelson and olsen ( 1978 ) , amemiya ( 1979 ) , heckman ( 1978 ) , and smith and blundell ( 1986 ) , with more efficient estimators proposed by newey ( 1987 ) and blundell and smith ( 1989 ) . on the contrary
, few studies have estimated censored quantile regression with endogenous regressors . while blundell and powell ( 2007 ) introduced control variables as in lee ( 2007 ) to deal with the endogeneity in censored quantile regression , their estimation method involved a high dimensional nonparametric estimation and can be computationally cumbersome .
chernozhukov _ et al_. ( 2014 ) also introduced control variables to account for endogeneity .
they proposed using quantile regression and distribution regression ( chernozhukov _ et al_. , 2013 ) to construct the control variables and extended the estimation method of chernozhukov and hong ( 2002 ) . in the bayesian framework , mean
regression models with endogenous variables have garnered a great deal of research attention from both the theoretical and the computational points of view ( _ e.g . _
rossi _ et al_. , 2005 ; hoogerheide _ et al_. , 2007a , 2007b ; conely _ et al_. , 2008 ; lopes and polson , 2014 ) .
however , despite the growing interest in and demand for bayesian quantile regression , the literature on bayesian quantile regression with endogenous variables remains sparse .
lancaster and jun ( 2010 ) utilised the exponentially tilted empirical likelihood and employed the moment conditions used in chernozhukov and hansen ( 2006 ) . in the spirit of lee ( 2007 ) , ogasawara and kobayashi ( 2015 ) employed a simple parametric model using two asymmetric laplace distributions for panel quantile regression
. however , these methods are only applicable to uncensored data .
furthermore , the model of ogasawara and kobayashi ( 2015 ) can be restrictive because of the shape limitation of the asymmetric laplace distribution , which can affect the estimates .
indeed , the modelling of the first stage error in this approach remains to be discussed .
based on the foregoing , this study proposes a flexible parametric bayesian endogenous tqr model .
the @xmath0-th quantile regression of interest is modelled parametrically following the usual bayesian quantile regression approach .
following lee ( 2007 ) , we introduce a control variable such that the conditional quantile of the error term is corrected to be zero and the parameters are correctly estimated . as in the approach of lee ( 2007 ) , the @xmath1-th quantile of the error term in the regression of the endogenous variable on the exogenous variables , which is often called the first stage regression , is also assumed to be zero .
we discuss the modelling approach for the first stage regression and consider a number of parametric and semiparametric models based on the extensions of ogasawara and kobayashi ( 2015 ) .
specifically , following wichitaksorn _ et al_. ( 2014 ) and naranjo _ et al_. ( 2015 ) , we employ the first stage regression models based on the asymmetric laplace distribution , skew normal distribution , and asymmetric exponential power distribution , for which the @xmath1-th quantile is always zero and is modelled by the regression function . to introduce more flexibility into the tail behaviour of the models based on the asymmetric laplace and skew normal distributions
, we also consider a semiparametric extension using the dirichlet process mixture of scale parameters as in kottas and krnjaji ( 2011 ) .
the value of @xmath1 is a priori unknown , while the choice of @xmath1 can affect the estimates . in this study , hence , @xmath1 is treated as a parameter to incorporate uncertainty and is estimated from the data .
the performance of the proposed models is demonstrated in a simulation study under various settings , which is a novel contribution of the present study .
we also illustrate the influence of the prior distributions on the posterior in the cases where valid and weak instruments are used .
the rest of this paper is organised as follows .
section [ sec : tobit ] introduces the standard bayesian tqr model with a motivating example .
then , section [ sec : approach ] proposes bayesian tqr models to deal with the endogenous variables .
the mcmc methods adopted to make inferences about the models are also described .
the simulation study under various settings is presented in section [ sec : sim ] .
the models are also illustrated by using the real data on the working hours of married women in section [ sec : real ] .
finally , we conclude in section [ sec : conc ] .
suppose that the response variables are observed according to @xmath3 then , consider the @xmath0-th quantile regression model for @xmath4 given by @xmath5 where @xmath6 is the vector of regressors , @xmath7 is the coefficient parameter , and @xmath8 is the error term whose @xmath0-th quantile is zero .
the @xmath0-th conditional quantile of @xmath9 is modelled as @xmath10 .
the equivariance under the monotone transformation @xmath11 of quantiles implies that the @xmath0-th conditional quantile of @xmath12 is given by @xmath13 the tqr model can be estimated by minimising the sum of asymmetrically weighted absolute errors @xmath14 where @xmath15 and @xmath16 denotes the indicator function ( powell , 1986 ) .
the bayesian approach assumes that @xmath17 follows the asymmetric laplace distribution , since minimising ( [ eqn : check ] ) is equivalent to maximising the likelihood function of the asymmetric laplace distribution ( koenker and machado , 1999 ; chernozhukov and hong , 2003 ) .
the probability density function of the asymmetric laplace distribution , denoted by @xmath18 , is given by @xmath19 where @xmath20 is the scale parameter and @xmath21 is the shape parameter ( yu and zhang , 2005 ) .
the mean and variance are given by @xmath22=\sigma\frac{1 - 2p}{p(1-p)}$ ] and @xmath23 .
the @xmath0-th quantile of this distribution is zero , @xmath24 .
assuming the prior distributions for the parameters , the parameters are estimated by using the mcmc method ( _ e.g .
_ yu and stander , 2007 ; kozumi and kobayashi , 2011 ) .
posterior consistency of bayesian quantile regression based on the asymmetric laplace distribution was shown by sriram _ et al_. ( 2013 ) .
estimates under the standard bayesian tqr model are biased when endogenous variables are included as regressors .
consider a simple motivating example where the dataset was generated from @xmath25 for @xmath26 , where @xmath27 , @xmath28 , @xmath29 and @xmath30.\ ] ] see also chernozhukov _ et al_. ( 2014 ) .
note that @xmath31 expresses the level of endogeneity .
while @xmath32 is an exogenous variable when @xmath33 , @xmath32 is endogenous when @xmath34 . since @xmath35 ,
the model can be rewritten as @xmath36 therefore , the standard model that models the conditional quantile of @xmath9 as @xmath37 produces biased estimates .
figure [ fig : fig1 ] shows the posterior distributions of @xmath38 , @xmath39 , and @xmath40 for the standard model for @xmath41 obtained by using the method of kozumi and kobayashi ( 2011 ) .
the vertical lines in the figure indicate the true values . in the case of @xmath33 ,
the posterior distributions are concentrated around the true values . however , in the case of @xmath42 , the posterior distributions are concentrated away from the true values . , @xmath39 , and @xmath40 using the standard bayesian tobit median regression ]
we propose the following model to deal with the endogenous variables : @xmath43 for @xmath44 , where @xmath6 is the vector of the exogenous variables whose the first element is @xmath45 , @xmath46 is the endogenous variable , @xmath47 , and @xmath48 is the exogenous variable not included in @xmath6 , which is also called the instrumental variable .
the term @xmath49 in ( [ eqn:2nd ] ) is called the control variable and is introduced to account for endogeneity .
note that @xmath50 indicates @xmath46 is endogenous .
we refer to ( [ eqn:1st ] ) as the first stage regression and to ( [ eqn:2nd ] ) as the second stage regression .
a similar form is found in lopes and polson ( 2014 ) in the context of the instrumental variable regression for means by using the cholesky - based prior .
following lee ( 2007 ) , the error term @xmath8 of the standard bayesian tqr is decomposed into the terms @xmath51 and @xmath52 .
it is assumed that relationship ( [ eqn:1st ] ) is specified correctly and the quantile independence of @xmath52 on @xmath53 conditional on @xmath54 : @xmath55 as in lee ( 2007 ) , we also assume @xmath56 where the @xmath1-th conditional quantile of @xmath54 is zero for some @xmath57 . we are mainly concerned with modelling the first stage error that satisfies ( [ eqn : alpha ] ) .
a simple and convenient approach is to assume @xmath58 as in ogasawara and kobayashi ( 2015 ) , since ( [ eqn : alpha ] ) is always satisfied for the asymmetric laplace distribution .
however , the asymmetric laplace distribution has limitations , such as peaky density , restrictive tail behaviour , and skewness .
when a model lacks fit to the data , the estimate of the conditional quantile would be away from the value such that ( [ eqn : alpha ] ) truly holds .
then , assuming @xmath54 is homoskedastic , the estimate of the intercept , @xmath59 , may be biased as well .
consequently , the estimate of @xmath60 would be affected through the introduced term @xmath51 .
when @xmath54 is heteroskedastic , the entire coefficient vector would be affected .
therefore , we consider some alternative models for the first stage error distribution .
recently , wichitaksorn _ et al_. ( 2014 ) considered a class of parametric distributions with a quantile constraint of the form ( [ eqn : alpha ] ) , including the asymmetric laplace distribution , and applied them in the context of quantile modelling .
furthermore , zhu and zinde - walsh ( 2009 ) , zhu and galbraith ( 2011 ) , and naranjo _ et al_. ( 2015 ) considered a flexible parametric distribution with the quantile constraint . based on these studies , we also consider the following two distributions to model the first stage error .
first , we consider the skew normal distribution denoted by @xmath61 , where @xmath62 is the scale parameter and @xmath57 is the shape parameter .
the probability density function is given by @xmath63 when @xmath64 , the distribution reduces to @xmath65 .
the mean and variance are given by @xmath66=\sqrt{\frac{\phi}{2\pi}}\frac{1 - 2\alpha}{\alpha(1-\alpha)}$ ] and @xmath67 ( see wichitaksorn _ et al_. , 2014 ) .
when the actual error distribution is close to the normal distribution , this distribution would lead to better performance than the asymmetric laplace distribution .
however , just as the asymmetric laplace distribution , the skewness and the quantile level of the mode are controlled by the single parameter @xmath1 .
second , we consider the asymmetric exponential power distribution treated by zhu and zinde - walsh ( 2009 ) , zhu and galbraith ( 2011 ) , and naranjo _ et al_. ( 2015 ) .
the probability density function of the asymmetric exponential power distribution , denoted by @xmath68 , is given by @xmath69 where @xmath62 is the scale parameter , @xmath57 is the skewness parameter , @xmath70 is the shape parameter for the left tail , and @xmath71 is the shape parameter for the right tail . after some reparameterisation
, the distribution reduces to the asymmetric laplace distribution when @xmath72 and to the skew normal distribution when @xmath73 .
the tails of the asymmetric exponential power distribution are controlled separately by @xmath74 and @xmath75 , respectively , and the overall skewness is controlled by @xmath1 . although the distribution is more flexible than the above two distributions , the posterior computation using mcmc would be inefficient , because it includes two additional shape parameters and it has no convenient mixture representation , apart from the mixture of uniforms that is inefficient , to facilitate an efficient mcmc algorithm .
the computational efficiency is also compared in section [ sec : sim ] .
in addition to the three parametric models , we also consider the semiparametric extension of the models based on the asymmetric laplace and skew normal distributions to achieve both flexibility and computational efficiency .
more specifically , the following two models using the dirichlet process mixtures of scales are considered : @xmath76 where @xmath77 denotes the dirichlet process with the precision parameter @xmath78 and the base measure @xmath79 .
for both models , we set @xmath80 as it is computationally convenient .
while those mixture models have the same limitation as the parametric versions in terms of skewness , they extend the tail behaviour of the error distribution preserving ( [ eqn : alpha ] ) ( kottas and krnjaji , 2009 ) .
hereafter , the models with the asymmetric laplace , skew normal , and asymmetric exponential power first stage errors are respectively denoted by al , sn , and aep , and those with the dirichlet process mixtures are denoted by aldp and sndp .
we must take care when selecting the @xmath1 value in ( [ eqn : alpha ] ) , as it is a part of the model specification and can thus affect the estimates ( lee , 2007 ) .
we treat @xmath1 as a parameter and estimate its value along with the other parameters . since @xmath1 determines the quantile level of the mode for all models considered here , our approach to modelling the first stage regression can also be regarded as a kind of mode regression ( see wichitaksorn _ et al_. , 2014 ) . to gain further flexibility , we might extend the model through a fully nonparametric mixture . several semiparametric models in the context of bayesian quantile regression with exogenous variables
have been proposed by kottas and gelfand ( 2001 ) , kottas and krnjaji ( 2009 ) , and reich _ et al_. ( 2010 ) .
for example , kottas and krnjaji ( 2009 ) considered the nonparametric mixture of uniform distributions for any unimodal density on the real line with the quantile restriction at the mode using the dirichlet process mixture ( see also kottas and gelfand , 2001 ) . in the more flexible model proposed by reich
_ et al_. ( 2010 ) , the mode of the error distribution does not have to coincide with zero .
this is achieved by using a nonparametric mixture of the quantile - restricted two - component mixtures of normal distributions .
however , their approaches are not directly applicable in the present context where the value of @xmath1 is estimated . if we were to estimate the quantile level for which the quantile restriction holds , the computation under the former model is expected to be extremely inefficient and unstable as the model involves many indicator functions , and @xmath1 and the intercept would be highly correlated .
the intercept would not be identifiable in the latter model .
we could further extend the model to account for heteroskedasticity such that @xmath81 for @xmath44 , where @xmath82 for all @xmath83 and the first element of @xmath84 is fixed to one ( _ e.g .
_ reich , 2010 ) . in this case , the @xmath1-th quantile of @xmath32 is given by @xmath85 as in the usual quantile regression .
however , since the first stage regression model is built based on ( [ eqn : alpha ] ) , models ( [ eqn:1st ] ) and ( [ eqn : hetero ] ) would produce identical estimates .
we next turn to the model of the new second stage error , @xmath52 , in ( [ eqn:2nd ] ) . since the @xmath0-th conditional quantile of @xmath52 is now zero
, we assume that @xmath86 as in the standard bayesian quantile regression approach .
we utilise the location scale mixture of normals representation for the asymmetric laplace distribution to facilitate an efficient mcmc method following kozumi and kobayashi ( 2011 ) ( see also kotz _ et al_. , 2001 ) .
the model is expressed in the hierarchical form given by @xmath87 for @xmath44 , where @xmath88 , @xmath89 , @xmath90 denotes the exponential distribution with mean @xmath91 , and @xmath92 the coefficient parameter @xmath93 is common to all first stage regression specifications .
first , we assume the normal prior for @xmath93 , since it is computationally convenient for the al , sn , aldp , and sndp models .
since we do not have information on the coefficient values , the variances are set such that the prior distributions are relatively diffuse .
our default choice is @xmath94 . for the scale parameters , @xmath95 for the al , sn , and aep distributions ,
a relatively diffuse inverse gamma distribution is assumed and the default choice is set to @xmath96 . for aep ,
we assume @xmath97 , where @xmath98 denotes the normal distribution with the mean @xmath99 and variance @xmath100 truncated on the interval @xmath101 .
a similar prior specification is found in naranjo _ et al_. ( 2015 ) . for all models , @xmath102 is assumed . for the semiparametric models , we need to specify the parameters of the inverse gamma base measure .
assuming that the data have been rescaled , @xmath103 and @xmath104 are chosen such that the variance of @xmath54 takes values between @xmath105 and @xmath106 with high probability ( _ e.g .
_ ishwaran and james , 2002 ) .
our default choice is @xmath107 and @xmath108 for aldp and @xmath109 for sndp . under this choice ,
when @xmath64 for aldp , @xmath110 as @xmath111 .
similarly , when @xmath112 , @xmath113 . for sndp , @xmath114 when @xmath64 and @xmath115 when @xmath112 .
for the precision parameter of the dirichlet process , @xmath116 , we assume @xmath117 such that both small and large values for @xmath116 , hence the number of clusters , are allowed . for the coefficient parameters in the second stage , @xmath7 and @xmath118
, we also assume relatively diffuse normal distributions .
our default choice of prior is @xmath119 .
similar to @xmath95 in the parametric first stage , we assume an inverse gamma prior for the scale of the al pseudo likelihood .
our default choice is @xmath96 .
the parameter @xmath120 accounts for the endogeneity and we need to take care in prior elicitation . when the data follow the bivariate normal distribution , as in the motivating example ( [ eqn : example ] ) , @xmath120 is equal to @xmath121 , where @xmath31 is the correlation coefficient and @xmath122 and @xmath123 are the standard deviations of the first and second stage errors , respectively . in this case , we may follow lopes and polson ( 2014 ) to determine the variance of the normal prior implied from an inverse wishart prior for the covariance matrix . however , we do not limit ourselves to normal data as the quantile regression approach is suitable for heteroskedastic and non - normal data , and the non - normal models are used in the first stage . in the literature on bayesian non - normal selection models , the prior distribution of @xmath120 is normal typically with a very small variance , such as @xmath124 ( _ e.g . _
munkin and trivedi , 2003 , 2008 ; deb _ et al_. , 2006 ) .
on the other hand , we use a more diffused prior to reflect our ignorance about @xmath120 and set our default choice of prior to be @xmath125 . when the instrument is weak , it is expected that our quantile regression models face the problem of prior sensitivity and that the posterior distributions exhibit sharp behaviour , as in the case of the bayesian instrumental variable regression model .
section [ sec : sim ] considers the alternative choices of the hyperparameters to study the prior sensitivity .
the proposed models are estimated by using the mcmc method based on the gibbs sampler .
we describe the gibbs sampler for the semiparametric models with aldp and sndp , which is an extension of the gibbs sampler described in kozumi and kobayashi ( 2011 ) and ogasawara and kobayashi ( 2015 ) .
the algorithms for the al and sn models can be obtained straightforwardly .
we also mention the algorithm for the aep model .
the variables involved in the dirichlet process are sampled by using the retrospective sampler ( papaspiliopoulos and roberts , 2008 ) and the slice sampler ( walker , 2007 ) .
first , we introduce @xmath126 and @xmath127 , such that @xmath128 .
then , as in walker ( 2007 ) , the gibbs sampler is constructed by working on the following joint densities @xmath129 where @xmath130 , @xmath131 , @xmath132 , and @xmath133 denotes the beta distribution with the parameters @xmath116 and @xmath134 ( sethuraman , 1994 ) .
we also let @xmath135 denote the minimum integer such that @xmath136 . for the aldp model , we utilise the mixture representation for the asymmetric laplace distribution to sample @xmath93 efficiently such that @xmath137 , @xmath138 , @xmath44 , where @xmath139 and @xmath140 are defined as in ( [ eqn : theta ] ) .
let us denote @xmath89 and @xmath141 .
our gibbs sampler proceeds by alternately sampling @xmath142 , @xmath143 , @xmath144 , @xmath145 , @xmath116 , @xmath93 , @xmath146 , @xmath1 , @xmath147 , @xmath148 , @xmath91 , and @xmath149 . * * sampling @xmath142 : * generate @xmath150 from @xmath151 for @xmath44 . *
* sampling @xmath143 : * generate @xmath152 from @xmath153 where @xmath154 for @xmath155 . *
* sampling @xmath144 : * generate @xmath156 from the multinomial distribution with probabilities @xmath157 for @xmath44 . *
* sampling @xmath145 : * generate @xmath158 from @xmath159 where @xmath160 + d_0.\ ] ] * * sampling @xmath116 : * assuming the gamma prior , @xmath161 , we use the method described by escobar and west ( 1995 ) to sample @xmath116 . by introducing @xmath162 , the full conditional distribution of @xmath116 is the mixture of two gamma distributions given by @xmath163 where @xmath164 is the number of distinct clusters and @xmath165 . *
* sampling @xmath93 : * assuming @xmath166 , @xmath93 is sampled from @xmath167 where @xmath168^{-1},\\ { \mathbf{g}}_1&=&{\mathbf{g}}_1\left[\sum_{i=1}^n{\mathbf{z}}_i\left(-\frac{\eta_p(y_i^*-{\mathbf{x}}_i'{\text{\boldmath{$\beta$}}}_p - \eta_pd_i-\theta_pg_i)}{\tau_p^2\sigma g_i}+\frac{d_i-\theta_\alpha h_i}{\tau_\alpha^2\phi_{k_i}h_i } \right)+{\mathbf{g}}_0^{-1}{\mathbf{g}}_0\right ] , \end{aligned}\ ] ] as the density of the full conditional distribution denoted by @xmath169 is given by @xmath170 * * sampling @xmath146 : * the full conditional distribution of @xmath171 is the generalised inverse gaussian distribution , denoted by @xmath172 .
the probability density function of @xmath173 is given by @xmath174 where @xmath175 is the modified bessel function of the third kind ( barndorff - nielsen and shephard , 2001 ) .
for @xmath44 , we sample @xmath171 from @xmath176 where @xmath177 *
* sampling @xmath1 : * the density of the full conditional distribution of @xmath1 is given by @xmath178 where @xmath179 and@xmath180 denote the full conditional and prior density of @xmath1 , respectively .
we use the random walk metropolis hastings ( mh ) algorithm to sample from this distribution . * * sampling @xmath147 : * the full conditional distribution of @xmath4 is given by @xmath181 * * sampling @xmath148 : * we sample @xmath89 in one block . assuming @xmath182 , the full conditional distribution is given by @xmath183 where @xmath184^{-1 } , \quad \tilde{{\mathbf{b}}}_1=\tilde{{\mathbf{b}}}_1\left[\sum_{i=1}^n\frac{\tilde{{\mathbf{x}}}_i(y_i^*-\theta_pg_i)}{\tau_p^2\sigma g_i}+ \tilde{{\mathbf{b}}}_0^{-1}\tilde{{\mathbf{b}}}_0\right].\ ] ] * * sampling @xmath91 : * assuming @xmath185 , we sample @xmath91 from @xmath186 where @xmath187 and @xmath188 . *
* sampling @xmath149 : * similar to @xmath171 , @xmath189 is sampled from @xmath190 where @xmath191 the gibbs sampler for sndp consists of sampling @xmath142 , @xmath143 , @xmath144 , @xmath145 , @xmath116 , @xmath93 , @xmath1 , @xmath147 , @xmath148 , @xmath91 , and @xmath149 .
the sampling algorithms for @xmath142 , @xmath143 , @xmath116 , @xmath147 , @xmath148 , @xmath91 , and @xmath149 remain the same as in the case of aldp .
the sampling scheme of @xmath144 and @xmath1 can be obtained by replacing @xmath192 with @xmath193 .
similar to the case of aldp , the density of the full conditional distribution is given by @xmath194 where @xmath195^{-1 } , \\
{ \mathbf{g}}_1({\text{\boldmath{$\gamma$}}})&=&{\mathbf{g}}_1({\text{\boldmath{$\gamma$}}})\left[\sum_{i=1}^n{\mathbf{z}}_i \left ( -\frac{\eta_p(y_i^*-{\mathbf{x}}_i'{\text{\boldmath{$\beta$}}}_p-\eta_pd_i-\theta_p g_i)}{\tau_p^2\sigma g_i}+\frac{4d_i(\alpha - i(d_i\leq{\mathbf{z}}_i'{\text{\boldmath{$\gamma$}}}))^2}{\phi_{k_i}}\right ) + { \mathbf{g}}_0^{-1}{\mathbf{g}}_0\right ] , \end{aligned}\ ] ] which is similar to the density of the normal distribution .
therefore , we sample @xmath93 by using the mh algorithm with the proposal distribution given by @xmath196
. since no convenient representation for the aep distribution is available , the full conditional distributions of the parameters in the first stage regression , @xmath93 , @xmath95 , @xmath1 , @xmath74 , and @xmath75 , are not in the standard forms .
therefore , we employ the adaptive random walk mh algorithm . although naranjo _ et al_. ( 2015 ) proposed the scale mixture of uniform representation for the aep distribution , the algorithm based on this representation would be inefficient , because it consists of sampling from a series of distributions that are truncated on some intervals such that the mixture representation holds and such intervals move quite slowly as sampling proceeds ( see also kobayashi , 2015 ) .
since the additional shape parameters in aep free up the role of @xmath1 , @xmath1 controls the overall skewness by allocating the weights on the left and right sides of the mode .
hence , the mcmc sample would exhibit relatively high correlation between @xmath1 and @xmath59 .
the models considered in the previous section are demonstrated using simulated data .
the aims of this section are ( 1 ) to compare the performance of the proposed models ( section [ sec : default ] ) , ( 2 ) to study the sensitivity to the prior settings , and ( 3 ) to illustrate the behaviour of the posterior distribution when the instrument is weak ( section [ sec : alt ] ) .
the data are generated from the model given by @xmath197 for @xmath26 , where @xmath198 assuming that a valid instrument is available , @xmath199 , @xmath200 , and @xmath201 .
the performance of the models is compared by considering the various settings for @xmath54 , while the distributions of @xmath52 are kept relatively simple in order that the true values of the quantile regression coefficients are tractable .
the following five settings are considered : + * setting 1 * @xmath202 , @xmath203 , + * setting 2 * @xmath204 , @xmath205 , + * setting 3 * @xmath206 , @xmath205 , + * setting 4 * @xmath207 , @xmath203 , + * setting 5 * @xmath208 , @xmath205 , + where @xmath209 denotes the skew @xmath210 distribution with the location parameter @xmath99 , scale parameter @xmath100 , skewness parameter @xmath211 , @xmath212 , and degree of freedom @xmath213 ( see azzalini and capitanio , 2003 ; fr@xmath214hwirth - schnatter and pyne , 2010 ) , and we set @xmath215 . in setting 1 ,
the error terms follow the bivariate normal distribution as in the motivating example in section [ sec : tobit ] .
setting 2 considers the fat tailed first stage regression . setting 3
considers a more difficult situation where the first stage error is fat tailed and skewed .
setting 4 replaces the first stage error of setting 1 with the heteroskedastic error with respect to the instrument .
setting 5 is also a challenging situation where the first stage error is fat tailed , skewed , and heteroskedastic . in settings 3 and 5 , the location parameters of the first stage error distributions are set such that the mode of @xmath54 is zero and the quantile level of the mode is @xmath216 .
the average censoring rates for the settings are around @xmath217 . for each setting , the data are replicated @xmath218 times .
we first estimated the proposed models under the default prior specifications ( see section [ sec : prior ] ) for @xmath219 and @xmath220 by running the mcmc for @xmath221 iterations and discarding the first @xmath222 draws as the burn - in period .
the standard bayesian tqr model was also estimated .
the bias and root mean squared error ( rmse ) of the parameters were computed over the @xmath218 replications . to assess the efficiency of the mcmc algorithm
, we also recorded the inefficiency factor , which was defined as a ratio of the numerical variance of the sample mean of the markov chain to the variance of the independence draws ( chib , 2001 ) .
table [ tab : default ] presents the biases , rmses , and median inefficiency factors for the parameters over the @xmath218 replications .
first , we examined the inefficiency factors .
overall , our sampling algorithms appear to be efficient , especially for al , sn , aldp , and sndp .
the table shows that the inefficiency factors for al , sn , aldp , and sndp are reasonably small for @xmath223 , @xmath118 , @xmath120 , @xmath224 , and @xmath225 . since @xmath1 and @xmath59 determine the quantile level of the mode and location of the mode , respectively , the mcmc sample exhibits correlation between @xmath1 and @xmath59 and this results in higher inefficiency factors for them .
hence , the inefficiency factors for @xmath226 tend to be higher than those for the other parameters .
this pattern is more profound in the case of aep where the inefficiency factors for @xmath1 , @xmath59 , and @xmath226 are quite high . since the additional shape parameters in aep free up the role of @xmath1 , the mcmc sample exhibits higher correlation between @xmath1 and @xmath59 .
furthermore , the inefficiency factors for the other parameters for aep are also higher than those for the other endogenous models .
next , we turn to the performance of the models .
as expected , tqr produces biased estimates in all cases .
the rmses for the proposed endogenous models are generally larger for @xmath219 , which is below the censoring point , than for @xmath41 .
the al and aldp models result in similar performance .
the aep model shows the largest rmses for @xmath59 and @xmath226 among the proposed models for all cases .
combined with the high inefficiency factors for those parameters , the convergence of the mcmc algorithm for aep may be difficult to ensure in the given simulation setting .
this finding suggests a considerable practical limitation and , thus , aep will not be considered henceforth .
the same limitation applies to the potentially more flexible nonparametric models discussed in section [ sec:1st ] .
table [ tab : default ] also shows that the estimation of the first stage regression can influence the second stage parameters .
for example , in setting 1 , the rmses for @xmath59 for sn and sndp are smaller than those for al and aldp , as the true model is the normal and thus sn and sndp produce smaller rmses for @xmath226 . similarly , in setting 4 ,
the rmses for @xmath226 for sn and sndp are smaller than those for al and aldp .
in addition , the heteroskedasticity in the first stage influences the performance of the slope parameters , resulting in slightly smaller rmses for @xmath223 for sn and sndp than for al and aldp .
however , the performance of the sn model becomes worse when the first stage error is fat tailed , since the skew normal distribution can not accommodate a fat tailed distribution .
while the results in setting 2 are somewhat comparable across the models , the table shows that sn results in larger biases and rmses in setting 3 and , especially , setting 5 . in setting 3 ,
sn results in larger rmses for @xmath226 than for al , aldp , and sndp . in setting 5 ,
given the heteroskedasticity of the first stage , the biases and rmses for the intercept and slope parameters for sn are larger than those for al , aldp , and sndp .
on the other hand , compared with sn , the semiparametric sndp model is able to cope with fat tailed errors and this produces results comparable with those for al and aldp . while the models result in reasonable overall performance , the results for settings 3 and 5 also illustrate the limitation of our modelling approach to some extent . in setting 3 , the models exhibit some bias in @xmath226 because of the lack of fit in the first stage .
this lack of fit , which is represented by the bias for @xmath59 , is reflected in the bias for @xmath226 .
the entire coefficient vector may be influenced by this lack of fit in the first stage in the presence of heteroskedasticity as in setting 5 .
the lack of fit in the first stage is also indicated by the biases in @xmath1 .
this finding implies that an inflexible first stage model can fail to estimate the true quantile such that ( [ eqn : alpha ] ) holds and that choosing the value of @xmath1 a priori could lead to biased estimates ( see the discussion in section [ sec:1st ] ) . for comparison purposes ,
we consider two alternative specifications for the inverse gamma base measure for the semiparametric models .
the following slightly less diffuse settings than the default are considered . for aldp , we consider @xmath227 such that @xmath228 and @xmath229 such that @xmath230 when @xmath64 . for sndp
, we consider @xmath231 such that @xmath232 and @xmath233 such that @xmath234 . for the other parameters
, we use the default prior specifications .
table [ tab : base ] presents the biases and rmses for aldp and sndp under the alternative base measures for @xmath219 and @xmath41 .
the results in table [ tab : base ] are essentially identical to those in table [ tab : default ] , suggesting that the default choice of the base measures provides reasonable performance .
next , the two alternative prior specifications for @xmath120 , @xmath91 , and @xmath95 are considered to study the prior sensitivity .
the first alternative specification considers the more diffuse priors given by @xmath235 , @xmath236 , and @xmath237 . the second alternative specification is the even more diffuse setting given by @xmath238 , @xmath239 , and @xmath240 .
for aldp and sndp , the default base measures are used . for @xmath7 , @xmath118 , and @xmath93
, we use the default specification .
table [ tab : prior ] presents the biases and rmses for al , sn , aldp , and sndp under the five simulation settings for @xmath219 and @xmath220 , showing that the result is robust with respect to the choice of hyperparameters .
we also considered some different prior choices for @xmath241 and @xmath93 , and obtained robust results .
these findings thus confirm the robustness of the results with respect to the choice of base measures and prior distributions provided that a valid instrument is available . in the context of mean regression models ,
however , when the instrument is weak , the posterior distribution is known to exhibit sharp behaviour in the vicinity of non - identifiability ( hoogerheide _ et al_. , 2007b ) and the posterior distribution is greatly affected by the prior specification ( _ e.g . _ lopes and polson , 2014 ) . here
, we illustrate the behaviour of the posterior distribution by using a weak instrument .
the data are generated from ( [ eqn : sim1 ] ) without the regressor : @xmath242 for @xmath26 , where @xmath243 , @xmath244 , @xmath245 , @xmath202 , and @xmath203 .
the al and sn models are estimated for @xmath219 by running the mcmc for 20000 iterations and discarding the first 5000 draws as the burn - in period under the three prior specifications previously considered .
figure [ fig : fig2 ] presents the joint posterior distribution of @xmath246 and @xmath247 for al and sn under the three prior specifications and shows that the posterior distribution is greatly affected by the prior specification .
the posterior distribution of @xmath40 becomes more diffuse as @xmath248 approaches zero .
this trend becomes more profound as we use more diffuse prior distributions , producing star shapes .
the figure also suggests that the prior distribution can act as an informative prior about the linear relationship between @xmath40 and @xmath249 .
similar results were also obtained under different prior specifications for @xmath7 , @xmath118 , and @xmath93 as well as for aldp and sndp . and
@xmath247 for al ( top row ) and sn ( bottom row ) ]
the proposed endogenous models are applied to the dataset on the labour supply of married women of mroz ( 1987 ) .
the dataset includes observations on @xmath250 individuals .
the response variable is the total number of hours in every @xmath218 hours the wife worked for a wage outside the home during 1975 . in the data , 325 of the 753 women worked zero hours and the corresponding responses
are treated as left censored at zero .
hence , the censoring rate is approximately @xmath251 .
the regressors of our model include years of education ( _ educ _ ) , years of experience ( _ exper _ ) and its square ( _ expersq _ ) , age of the wife ( _ age _ ) , number of children under 6 years old ( _ kidslt6 _ ) , number of children equal to or greater than 6 years old ( _ kidsge6 _ ) , and non - wife household income ( _ nwifeinc _ ) .
we treat _ nwifeinc _ as an endogenous variable because it may be correlated with the unobserved household preference for the labour force participation of the wife . as an instrument
, we include the years of education of the husband ( _ huseduc _ ) , since this can influence both his income and the non - wife household income , but it should not influence the decision of the wife to participate in the labour force .
smith and blundell ( 1986 ) considered a similar setting where non - wife income was considered to be endogenous and the education of the husband was employed as the instrumental variable .
they applied the endogenous tobit model to data derived from the 1981 family expenditure survey in the united kingdom . using the default prior specifications ,
the aldp and sndp models are estimated for @xmath252 by running the mcmc for @xmath253 iterations and discarding the first @xmath254 draws as the burn - in period .
convergence is monitored by using the trace plots and gelman - rubin statistic for two chains with widespread starting values ( gelman _ et al_. , 2014 ) .
the upper bounds of the gelman - rubin confidence intervals for the selected parameters , @xmath255 , @xmath118 , @xmath120 , @xmath256 , @xmath257 , and @xmath1 , for sndp in the case of @xmath219 are @xmath258 , @xmath258 , @xmath258 , @xmath259 , @xmath259 , and @xmath260 , respectively .
figure [ fig : fig3 ] presents the post burn - in trace plots for these parameters and shows the evidence of convergence of the chains . ]
first , we present the results for the representative quantiles , @xmath219 , @xmath220 , and @xmath261 .
table [ tab : real ] shows the posterior means , 95% credible intervals , and inefficiency factors for aldp and sndp for these quantiles .
the table shows that the sampling algorithm worked efficiently as the inefficiency factors are reasonably small .
the posterior means for the instrument , _ huseduc _ , are positive and the 95% credible intervals do not include zero for all cases for both models , implying that _ huseduc _ is a valid instrument . for @xmath41 ,
the posterior means for @xmath120 are @xmath262 and @xmath263 for aldp and sndp , respectively , and the 95% credible intervals do not include zero .
therefore , it is suggested that non - wife income be treated as an endogenous variable for the median regression . to study the endogeneity in non - wife household income across quantiles , the posterior distributions of @xmath120 are presented .
the results across the quantiles can be best understood by plotting the posterior distributions as a function of @xmath0 .
figure [ fig : fig4 ] shows the posterior means and 95% credible intervals of @xmath120 for aldp and sndp for @xmath252 .
the figure shows that the two models produced similar results and that the posterior distributions of @xmath120 are concentrated away from zero for the mid quantiles .
specifically , for @xmath264 , the 95% credible intervals do not include zero for either model .
there are notable peaks around @xmath265 , where the posterior means of @xmath120 under the default prior specifications are @xmath266 and @xmath267 with the 95% credible intervals @xmath268 and @xmath269 for aldp and sndp , respectively .
this is an interesting result considering that the censoring rate is @xmath251 .
the result implies that the effect of the endogeneity of non - wife income is the most profound when the wife is about to decide whether to enter the labour force .
when the opportunity cost of labour supply is very high ( lower quantile ) or the wife works on a more regular basis ( higher quantile ) , such endogeneity diminishes .
smith and blundell ( 1986 ) also reported that non - wife income is endogenous by using the endogenous tobit regression model .
the mean of our dataset is @xmath270 , which approximately corresponds to the @xmath271-th quantile . for @xmath272 , the posterior mean of @xmath120 for aldp is @xmath273 with the 95% credible interval @xmath274 and that for sndp is @xmath275 with the 95% credible interval @xmath276 .
the figure also shows the posterior means and 95% credible intervals under the two alternative prior specifications considered in section [ sec : alt ] , confirming that our results are robust with respect to the prior specifications . under the default and alternative priors for @xmath252 ]
figure [ fig : fig5 ] compares the posterior means and 95% credible intervals of @xmath241 for sndp , aldp , and tqr for @xmath252 .
the results for sndp and aldp are quite similar .
the figure clearly shows that the posterior distributions for the key variable , _ nwifeinc _
, for the proposed models and tqr exhibit some differences for @xmath264 , where _ nwifeinc _ is indicated to be endogenous .
the difference becomes the most profound around @xmath265 for which the posterior mean for _ nwifeinc _ is @xmath278 for aldp , @xmath279 for sndp , and @xmath280 for tqr , implying a stronger effect of non - wife income when endogeneity is taken into account .
the posterior distributions for _ nwifeinc _ for aldp and sndp are more dispersed than that for tqr for all @xmath0 . while the 95% credible intervals include zero for all models for the upper quantiles , for the lower quantiles , such as @xmath219 , those for aldp and sndp include zero and those for tqr do not .
differences in the results are also observed for other variables . for @xmath265 ,
the posterior means for _ educ _ and _ age _ are respectively @xmath281 and @xmath282 for aldp , @xmath283 and @xmath284 for sndp , and @xmath285 and @xmath286 for tqr . for the upper quantiles , @xmath287 , the 95% credible intervals for _ educ _ include zero for the proposed models , while those for tqr do not , implying that an additional year of education does not increase the working hours for those quantiles when the endogeneity from non - wife income is taken into account . for _ expersq
_ , the endogenous models result in slightly more dispersed posterior distributions for @xmath288 .
the posterior means for @xmath265 are @xmath289 , @xmath290 , and @xmath291 for aldp , sndp , and tqr , respectively . for _
kidsge6 _ , the posterior means for @xmath265 are @xmath292 , @xmath293 , and @xmath294 for aldp , sndp , and tqr , respectively
. however , the 95% credible intervals include zero for all @xmath0 for all models .
on the other hand , the figure also shows that the models produced similar results for _ exper _ and _ kidslt6 _ for all @xmath0 . for aldp , sndp , and tqr for @xmath252 ]
we proposed bayesian endogenous tqr models using parametric and semiparametric first stage regression models built around the zero @xmath1-th quantile assumption .
the value of @xmath1 determines the quantile level of the mode of the error distribution and is estimated from the data . from the simulation study ,
the al , aldp , and sndp models worked relatively well for the various situations , while they faced the same limitation pointed out by kottas and krnjaji ( 2011 ) . on the other hand
, the sn model could not accommodate the fat tailed first stage errors .
although aep could be a promising model in terms of flexibility , the inefficiency of the mcmc algorithm largely limits its applicability in practice .
the development of a more convenient mixture representation for the aep distribution is thus required . from application to data on the labour supply of married women , the effect of the endogeneity in non - wife income was found to be the most profound for the quantile level close to the censoring rate . for this quantile ,
some differences in the parameter estimates between the endogenous and standard models were found , such as the stronger effect of non - wife income on working hours .
this study only considered the case of continuous endogenous variables .
we are also interested in incorporating endogenous binary variables into a bayesian quantile regression model .
an important extension might therefore be addressing multiple endogenous dummy variables to represent selection among multiple alternatives , such as the choice of a hospital and insurance plan , as considered in geweke _ et al_. ( 2003 ) and deb _ et al_. ( 2006 ) .
however , such an extension would be challenging with respect to the assumptions that must be imposed on the multivariate error terms .
we leave these issues to future research .
the author would like to thank the seminar participants at the second baysm and esobe 2014 and the anonymous referees for the valuable comments to improve the manuscript .
the computational results were obtained by using ox version 6.21 ( doornik , 2007 ) .
this study was supported by jsps kakenhi grant numbers 25245035 , 26245028 , 26380266 , and 15k17036 .
azzalini , a. and capitanio , a. ( 2003 ) .
`` distributions generated by perturbation of symmetry with emphasis on a multivariate skew @xmath210-distribution , '' _ journal of royal statistical society series b _ , * 65 * , 367389 .
barndorff - nielsen , o. e. and shephard , n. ( 2001 ) .
`` non - gaussian ornstein - uhlenbeck - based models and some of their uses in financial econometrics , '' _ journal of royal statistical society series b _ , * 63 * , 167241 .
deb , p. , munkin , k. m. , and trivedi , k. p. ( 2006 ) .
`` bayesian analysis of the two - part model with endogeneity : application to health care expenditure , '' _ journal of applied econometrics _ , * 21 * , 10811099 .
fr@xmath214hwirth - schnatter , s. and pyne , d. ( 2010 ) .
`` bayesian inference for finite mixtures of univariate and multivariate skew - normal and skew-@xmath210 distributions , '' _ biostatistics _ , * 11 * , 317336 .
hoogerheide , l. f. , kleibergen , f. , van dijk , h. k. ( 2007a ) .
`` natural conjugate priors for the instrumental variables regression model applied to the angrist - krueger data , '' _ journal of econometrics_**138 * * , 63103 .
hoogerheide , l. , kaashoek , j. f. , and van dijk , h. k. ( 2007b ) .
`` on the shape of posterior densities and credible sets in instrumental variable regression models with reduced rank : an application of flexible sampling methods using neural networks , '' _ journal of econometrics _ , * 139 * , 154180 .
ishwaran , h. and james , l. f. ( 2002 ) . `` approximate dirichlet process computing in finite normal mixtures : smoothing and prior information , '' _ journal of computational and graphical statistics _ , * 11 * , 508532 .
kobayashi , g. ( 2015 ) .
`` skew exponential power stochastic volatility model for analysis of skewness , non - normal tails , quantiles and expectiles , '' _ computational statistics _ , doi:10.1007/s00180 - 015 - 0596 - 4 .
, kozubowski , t. j. , and podgrski , k. ( 2001 ) . _ the laplace distribution and generalizations : a revisit with applications to communications , economics , engineering , and finance _ , birkh@xmath295user , boston .
munkin , m. k. and trivedi , p. k. ( 2003 ) . `` bayesian analysis of a self - selection model with multiple outcomes using simulation - based estimation : an application to the demand for healthcare , '' _ journal of econometrics _
, * 114 * , 197220 .
ogasawara , k. and kobayashi , g. ( 2015 ) .
`` the impact of social workers on infant mortality in inter - war tokyo : bayesian dynamic panel quantile regression with endogenous variables , '' _ cliometrica _ , * 9 * , 97130 .
wichitaksorn , n. , choy , s. t. b. , and gerlach , r. ( 2014 ) .
`` a generalized class of skew distributions and associated robust quantile regression models , '' _ the canadian journal of statistics _ , * 42 * , 5779596
. zhu , d. and galbraith , j. w. ( 2011 ) . ``
modeling and forecasting expected shortfall with the generalized asymmetric student-@xmath210 and asymmetric exponential power distributions , '' _ journal of empirical finance _ , * 18 * , 765778 . | this study proposes @xmath0-th tobit quantile regression models with endogenous variables . in the first stage regression of the endogenous variable on the exogenous variables , the assumption that the @xmath1-th quantile of the error term is zero
is introduced .
then , the residual of this regression model is included in the @xmath0-th quantile regression model in such a way that the @xmath0-th conditional quantile of the new error term is zero .
the error distribution of the first stage regression is modelled around the zero @xmath1-th quantile assumption by using parametric and semiparametric approaches .
since the value of @xmath1 is a priori unknown , it is treated as an additional parameter and is estimated from the data .
the proposed models are then demonstrated by using simulated data and real data on the labour supply of married women . * keywords * : asymmetric laplace distribution ; bayesian tobit quantile regression ; dirichlet process mixture ; endogenous variable ; markov chain monte carlo ; skew normal distribution ; |
the exosphere is the upper layer of any planetary atmosphere : it is a quasi - collisionless medium where the particle trajectories are more dominated by gravity than by collisions . above the exobase , the lower limit of the exosphere ,
the knudsen number @xcite becomes large , collisions become scarce , the distribution function can not be considered as maxwellian anymore and , gradually , the trajectories of particles are essentially determined by the gravitation and radiation pressure by the sun .
the trajectories of particles , subject to the gravitational force , are completely solved with the equations of motion , but it is not the case with the radiation pressure @xcite . in the absence of radiation pressure , we can distinguish three types of trajectories for the exospheric particles : * the escaping particles come from the exobase and have a positive mechanical energy : they can escape from the gravitational influence of the planet with a velocity larger than the escape velocity .
these particles are responsible for the jeans escape @xcite .
they can also be defined as crossing only once the exobase , * the ballistic particles also come from the exobase but have a negative mechanical energy , they are gravitationally bound to the planet .
they reach a maximum altitude and fall down on the exobase if they do not undergo collisions .
they cross the exobase twice , * the satellite particles never cross the exobase .
they also have a negative mechanical energy but their periapsis is above the exobase : they orbit along an entire ellipse around the planet without crossing the exobase .
the satellite particles result in their major part from ballistic particles undergoing few collisions mainly near the exobase ( @xcite ) .
thus , they do not exist in a collisionless model of the exosphere .
the radiation pressure disturbs the conics ( ellipses or hyperbolas ) described by the particles under the influence of gravity .
the resonant scattering of solar photons leads to a total momentum transfer from the photon to the atom or molecule @xcite . in the non - relativistic case , assuming an isotropic reemission of the solar photon , this one is absorbed in the sun direction and scattered with the same probability in all directions . for a sufficient flux of photons in the absorption wavelength range
, the reemission in average does not induce any momentum transfer from the atom / molecule to the photon .
the momentum variation , each second , between before and after the scattering imparts a force , the radiation pressure .
@xcite proposed to analyze its effect on the structure of planetary exospheres . in particular
, they highlighted analytically the tail " phenomenon at earth : the density for atomic hydrogen , which is sensitive to the lyman-@xmath0 photons , is higher in the nightside direction than in the dayside direction in the earth corona .
nevertheless , their work was limited only to the sun - planet axis , with a null component assumed for the angular momentum around the sun - planet axis .
we thus generalize here their work to a full 3d calculation , in order to investigate the influence of the radiation pressure on the trajectories ( this paper ) , as well as the density profiles and escape flux ( following works ) .
this problem is similar to the so - called stark effect @xcite : the effect of a constant electric field on the atomic hydrogen s electron .
its study can be transposed to celestial mechanics in order to describe the orbits of artificial and natural satellites in the perturbed ( e.g. by the radiation pressure force ) two - body problem .
a recent description of the stark effect solutions was already given by @xcite .
however , they give the analytical solutions of all trajectories only in the 2d case ( and for bounded trajectories in the 3d case ) , and their formulas have some issues as will be discussed later .
another analytical study proposed by @xcite uses the weierstrassian formulations to solve the motions for bounded and unbounded trajectories and to find periodic motions .
also , the motion can be approached numerically by developing the equations of motion in taylor series but this leads to some issues for high eccentricities @xcite
. @xcite compared recently these methods and their computing efficiencies . in this paper , based on the same formalism as @xcite
, we provide for the first time the complete exact 3d solutions of the stark effect ( and its celestial mechanics analogue ) for any initial condition and for both bounded and unbounded trajectories .
the first section describes the formalism used , before the sections [ model]/[section3]/[time ] provide the equations of motion and time .
we then discuss about circular orbits in section [ circular ] , while a comparison with previous works is given in section 6 , before we conclude in section [ summary ] .
in this work , we decide to study the effect of the radiation pressure on atomic hydrogen in particular . nevertheless , this formalism can be applied to any species subject to this force or to the interplanetary dust .
we model the radiation pressure by a constant acceleration @xmath1 coming from the sun . according to @xcite , this acceleration depends on the line center solar lyman-@xmath0 flux @xmath2 , in @xmath3 [email protected]@xmath5.@xmath5 : @xmath6 in spherical coordinates
, the hamiltonian of one hydrogen atom can be written : @xmath7 with @xmath8 the distance from the planet , @xmath9 the solar angle , @xmath10 the angle with respect to the ecliptic plane , @xmath11 , @xmath12 and @xmath13 the conjugate momenta .
@xmath14 represents the gravitational potential and @xmath15 the potential energy from the radiation pressure acceleration @xmath1 .
an example of trajectory of a @xmath16 atom subject to the radiation pressure is given in the figure [ trajectoire ] .
this problem is similar to the classical stark effect @xcite : a constant electric field ( here the radiation pressure ) is applied to an electron ( here an hydrogen atom ) attached to a proton ( here the planet ) .
both systems are equivalent because the force applied by the proton ( the planet ) to the electron ( the hydrogen atom ) , i.e. the electrostatic force , varies in @xmath17 as the gravitational force from the planet on the hydrogen atom .
thus , we adopt the same formalism as @xcite and use the parabolic coordinates .
we use the transformation : @xmath18 with @xmath19 and @xmath20 always positive .
consequently , the hamiltonian becomes : @xmath21 independent from @xmath22 and @xmath10 . according to hamiltonian canonical relations
, we have : @xmath23 in this new system of coordinates , we study this hamiltonian .
first , @xmath24 is independent from @xmath22 explicitly .
@xmath24 is a conserved quantity along the time and corresponds to the mechanical energy @xmath25 of the system .
moreover , as @xmath24 is independent from @xmath10 , according to canonical relations : @xmath26 thus , @xmath27 is another constant of the motion .
once @xmath25 and @xmath13 defined , the equation [ hamiltonuw ] can be rewritten : @xmath28 the left hand side is a function dependent only on @xmath19 and @xmath29 , the right hand side depends only on @xmath20 and @xmath30
. as both functions are equal and independent , they are equal to a constant @xmath31 , a separation constant : @xmath32 the motion possesses three constants : @xmath25 , @xmath31 and @xmath13 .
the equation [ a ] allows to express @xmath29 ( respectively @xmath30 ) as a functions of @xmath25 , @xmath31 , @xmath13 and @xmath19 ( respectively @xmath20 ) : @xmath33 with @xmath34 we have already introduced the hamiltonian @xmath24 of the system .
we can extend the approach according to hamilton - jacobi equations : @xmath35 where @xmath36 is the hamilton s principal function or action .
this function depends on initial conditions ( as @xmath37 , @xmath38 , @xmath39 and @xmath40 ) and the actual position of the particle ( as @xmath19 , @xmath20 , @xmath10 and @xmath22 ) .
as previously demonstrated , @xmath41 and @xmath27 are constants .
thus , @xmath42 and leads to : @xmath43\ ] ] with @xmath44 the part of the action independent from @xmath40 , @xmath22 , @xmath39 and @xmath10 .
moreover , the action @xmath44 can be separated into two parts : one written as a function of @xmath19 and @xmath29 coordinates , the other one with @xmath20 and @xmath30 . by definition , according to the hamilton - jacobi equations , we have : @xmath45 with @xmath29 ( resp .
@xmath30 ) a function only of @xmath19 ( resp .
@xmath20 ) , assuming @xmath25 , @xmath31 and @xmath13 values already fixed by initial conditions .
then , we can separate again the action , leading to : @xmath46=\mathcal{s}_{u}[u , e , a , p_{\phi}]+\mathcal{s}_{w}[w , e , a , p_{\phi}]\ ] ] @xmath47 according to the equation [ pupw ] , we have the following relations : @xmath48 with @xmath49 @xmath50 and @xmath51 are effective potentials applied in @xmath19 and @xmath20 directions ( represented in the figure [ potentiel ] ) .
these potentials play key roles for the motion because they constrained the motion in @xmath19 and @xmath20 directions independently . for the motion of the particle , we must respect two conditions : @xmath52 and @xmath53 .
these conditions are analogous to : @xmath54 ( corresponding to @xmath55 , blue line ) and @xmath56 ( corresponding to @xmath57 , red line ) for a set of @xmath25 , @xmath31 and @xmath27 values .
the motion is possible only in the area where the potential is below the mechanical energy @xmath25 , represented by the black horizontal line .
the different roots of @xmath58 and @xmath59 are displayed and correspond to the intersection of the potentials with the horizontal black line .
@xmath60 is the dimensionless value referring to @xmath61 ( cf .
section [ dimensionless_section ] , table [ dimensionless ] ) .
notice that @xmath57 will cross only once the energy level @xmath62 for too high or too low @xmath62 values . ]
these both conditions are more restrictive than the usual @xmath63 where @xmath64 is the potential energy .
@xmath58 is a polynom of degree 3 with @xmath65 .
this polynom possesses three roots , whose one is real at least . as @xmath66 ,
one of these roots is real negative , according to intermediate value theorem since @xmath67 .
nevertheless , the motion occurs for positive @xmath19 values , and we know this motion exists .
it implies there is an interval in @xmath68 such as @xmath69 ( otherwise , there is no motion in @xmath19-direction , no possible physically , cf .
eq . [ conditions ] ) .
to comply with this last condition , both other roots are real and positive . in summary , @xmath58 has three real roots : one negative and two positive .
we call each root @xmath37 , @xmath61 and @xmath70 such as @xmath71 and the @xmath19-motion is restricted to @xmath72 $ ] ( @xmath73 $ ] in term of dimensionless quantities , as can be seen in the figure [ potentiel ] ) .
@xmath59 is a polynom of degree 3 with @xmath74 .
this polynom possesses three roots , whose one is real at least . as @xmath75 , one of these roots is real positive , according to intermediate value theorem since @xmath74 .
nevertheless , the motion occurs for positive @xmath20 values .
we have restrictions on both other roots : they must be both real positive , both real negative or both complex conjugates . in the case where the three roots are real positive
, we call each root @xmath38 , @xmath76 and @xmath77 such as @xmath78 and the motion is restricted to @xmath79\cup[w_{0};+\infty[$ ] ( as can be seen in the figure [ potentiel ] with the dimensionless quantities , cf .
section [ dimensionless_q ] ) . in the case with one positive root and both other complex or real negative
, we call each root @xmath38 ( the positive one ) , @xmath76 and @xmath77 ( keep the same order as previously defined if they are real ) such as the motion is restricted to @xmath80 only ( @xmath81 in term of dimensionless quantities )
. each constant value of @xmath19 or @xmath20 defines a paraboloid in three dimensions . for each interval , constrained by fixed values of @xmath19 and @xmath20
, the motion will be contained between the paraboloids defined by these limit values as shown in the figure [ restriction1 ] .
right panel : the red area corresponds to the allowed region for the w - motion .
the regions are limited by paraboloids ( parabolas here by projection ) .
the sun location is shown with a yellow circle.,title="fig:"]-motion .
right panel : the red area corresponds to the allowed region for the w - motion .
the regions are limited by paraboloids ( parabolas here by projection ) .
the sun location is shown with a yellow circle.,title="fig : " ] for the @xmath19-motion , this is always limited by two paraboloids defined by @xmath82 and @xmath83 as shown by the blue area in the figure [ restriction1 ] ( left panel ) .
similarly , for the @xmath20-motion , there are two cases : the motion is constrained between one paraboloid ( @xmath84 ) and the infinity or between two paraboloids ( @xmath85 and @xmath86 ) .
both cases are represented by the red area in the figure [ restriction1 ] ( right panel ) .
as previously demonstrated , the motion is constrained in specific areas of the 3d space .
the motion is always between two paraboloids due to restrictions on @xmath19 but it can be constrained between two other paraboloids or only one regarding @xmath20 .
thus , the motion is constrained by four paraboloids or three , opened to infinity as shown on the figure [ restriction2 ] by the green area . .
the final allowed region is in green .
right panel : representation in green of the trajectory in the figure [ trajectoire ] in the @xmath87 frame at earth.,title="fig : " ] .
the final allowed region is in green .
right panel : representation in green of the trajectory in the figure [ trajectoire ] in the @xmath87 frame at earth.,title="fig : " ] nevertheless , there is at this point no other information on the exact motion of the particle .
as shown in the figure [ restriction2 ] ( right panel ) , the particle seems to explore all the area when its motion is restricted by four paraboloids .
this is simply an observation .
how can we prove that without any information on the exact motion of the particle ? according to the poincar recurrence theorem , if the dynamical trajectory of the autonomous system evolves in a finite volume of the phase space , then in any domain , as small as it could be , there are at least two points which belong to the same trajectory . here , the motion is constrained in space with the four paraboloids but also in velocity because @xmath29 and @xmath30 are finite values and @xmath27 is constant .
all the positions in this part of the phase space can belong to the same trajectory .
this is also linked to the kam theory : here , the perturbation ( the radiation pressure ) affects the periodic motion ( the ellipse , bounded trajectories ) but it can remain quasi - periodic . as we will see , the global motion is not periodic but @xmath19 and @xmath20 motions possess their own period according to another parameter .
the global motion can be periodic only if all periods are commensurable .
this is an important result because if the particle belongs to an area crossing the exobase and if this area is closed in the phase space , along a finite time , the particle will again cross the exobase .
we can now define the ballistic and satellite particles as follows : both populations have no elliptic trajectories due to the radiation pressure , but they evolve in a closed domain . depending on their constants of the motion , we can easily determine whether they cross ( i.e. the domain crosses ) the exobase or not , corresponding to ballistic and satellite particles respectively . escaping particles
are in the case where the initial value of @xmath20 is higher than the highest real root of @xmath59 and their available area is opened to the infinity .
thus , the theorem can not be applied here .
even if the restriction area crosses the exobase , the particle may come from infinity , come close to the exobase and go away without crossing the exobase .
we need , in order to identify escaping particles ( that cross the exobase and go to the infinity ) , to track along the time the particle to know if these particles come from the exobase or not .
thus , it is necessary to solve their trajectory along the time . for convenience , as usual in fluid mechanics
, we define characteristic quantities .
we decide to write all equations with dimensionless parameters .
first , for distance , we define : @xmath88 this characteristic value was introduced by @xcite and defines the limit distance where the radiation pressure overwhelms the gravitation of the planet . then
, we rewrite : @xmath89 where @xmath90 and @xmath91 are the dimensionless quantities associated to @xmath19 and @xmath20 .
the energy @xmath25 is adimensionned with the use of the characteristic energy @xmath92 . for @xmath29 and @xmath30 , we express them in units of @xmath93 , whereas @xmath27 is expressed in units of @xmath94 .
finally , we choose @xmath95 as the unit for @xmath31 and the unit for the adimensionless time @xmath96 is derived from the units of @xmath97 and @xmath98 .
we summarize these changes in table [ dimensionless ] . ' '' '' & & + ' '' '' & & + ' '' '' @xmath19 @xmath20&@xmath99 & @xmath90 @xmath91 + ' '' '' @xmath29 @xmath30&@xmath100 & @xmath101 @xmath102 + ' '' '' @xmath13&@xmath103 & @xmath104 + ' '' '' @xmath25&@xmath92&@xmath62 + ' '' '' @xmath31&@xmath95&@xmath105 + ' '' '' @xmath50 @xmath51&@xmath92&@xmath55 @xmath57 + ' '' '' @xmath22&@xmath106&@xmath107 + we express all the previous equations as a function of these new quantities : @xmath108 @xmath109 @xmath110 @xmath111 with @xmath112 : @xmath113 the jeans parameter at the distance @xmath114 .
we can also reduce the equations of the motion .
we introduce the dimensionless time @xmath96 .
in this part , we give implicit expressions for the dynamical trajectories of the particles , under the influence of both gravity and radiation pressure .
such expressions were already given for the 2d case with @xmath116 and partially generalized to the 3d case in @xcite with jacobi elliptic functions ( only bounded ) @xcite and in @xcite with the weierstrass functions for bounded and unbounded trajectories .
this last work dealt with the dynamical trajectories of artificial satellites but it can be applied to exospheric species subject to the radiation pressure .
we propose here corrections as well as a better way to give simple " expressions for dynamical trajectories .
we also provide expressions for unbounded trajectories that are missing in the literature . in the same way
, we introduce our notations for the next papers to be published , where the influence of the radiation pressure on the density profiles and escape flux will be investigated . according to the previous part
, we have different restrictions on the motion and thus , we must distinguish the cases , that will correspond to the different types of possible trajectories .
we may thus define the ballistic / satellite / escaping populations based on the roots of the @xmath58 and @xmath59 polynoms .
the trajectories are not elliptic or hyperbolic at all but one can keep their basic definitions : crossing twice the exobase for ballistic particles , orbiting but not crossing the exobase for satellite particles , coming from the exobase and escaping to the infinity for escaping particles .
the number of real roots is a key parameter for the analytical resolution of the trajectories .
the previous equations of the motion , expressed as a function of time , must be rewritten as a function of a new variable @xmath117 ( the subscript 0 is necessary to distinguish from @xmath96 ) defined as : @xmath118 until we know the expression of @xmath119 and @xmath120 , we can not express solutions as a function of time .
the system of equations leads now to : @xmath121 all solutions , trajectories and time , are parametrized according to the variable @xmath117 without physical sense .
we solve the differential equation describing the @xmath19-motion : @xmath122 with @xmath123 , @xmath60 and @xmath124 being real roots .
first , we set @xmath125 .
the motion occurs always with @xmath126 and @xmath127 to justify this change .
we obtain : @xmath128 now , we set @xmath129 .
@xmath130 finally , we have : @xmath131 we define @xmath132 as : @xmath133 @xmath134 the final equation is : @xmath135 the solution of this equation is : @xmath136 @xmath137 is a jacobi elliptic function and @xmath138 depends on initial conditions . the final expression for @xmath90 is : @xmath139 with @xmath140 and @xmath141 other jacobi elliptic functions .
these functions are @xmath142-periodic ( see appendix , eq .
[ goodformule ] ) . to determine @xmath138
, we preferably choose the second expression for the sign .
the initial conditions give us @xmath143 , then : @xmath144 where @xmath145 is defined as the reciprocal function of @xmath141 : @xmath146 with @xmath147 the incomplete elliptic integral of the first kind ( see appendix , eq . [ goodformule ] ) .
the sign depends on @xmath148 : @xmath149 if @xmath148 is positive ( resp .
negative ) , @xmath90 increases ( resp . decreases ) and @xmath150 increases ( resp . decreases ) too .
then , @xmath138 is positive ( resp .
negative ) .
this solution corresponds to the solution @xmath151 given by @xcite .
the @xmath90-motion is always constrained .
the characteristics of the motion / particle ( ballistic , satellite or escaping ) is thus determined by conditions on @xmath152 . for the @xmath20-motion
, we need to solve the differential equation : @xmath153 @xmath154 is a positive real root but we must distinguish between various cases for the other roots : @xmath155 and @xmath156 may be real or complex conjugate roots .
we must consider two cases : @xmath157 and @xmath158 .
the second one occurs mathematically and not only physically when @xmath156 and @xmath155 are positive .
[ [ a - for - w_-ww _ ] ] a ) for @xmath159 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + this case occurs for bounded trajectories ( ballistic or satellite trajectory ) . here
, this is the same treatment as for the @xmath90-motion . after setting the two following substitutions @xmath160 and @xmath161
, we obtain a similar solution : @xmath162 the final expression for @xmath91 is @xmath163 the initial conditions giving us @xmath164 and then leads to : @xmath165 the sign depends on @xmath166 : @xmath167 if @xmath166 is positive ( resp .
negative ) , @xmath91 increases ( resp .
decreases ) and @xmath168 increases ( resp . decreases ) too .
then , @xmath169 is negative ( resp .
positive ) .
this solution corresponds to the solution @xmath170 given by @xcite .
[ [ b - for - ww_0 ] ] b ) for @xmath157 + + + + + + + + + + + + + + + + + + + + + + + + + + + + first , the motion occurs only for @xmath171 ( corresponding to escaping particles ) .
then , we set @xmath172 . we obtain : @xmath173 now , we set @xmath161 . finally , we have : @xmath174 we define @xmath175 as : @xmath176 @xmath177 the solution of this equation is @xmath178 @xmath179 is a jacobi elliptic function and @xmath169 depends on the initial conditions .
the final expression for @xmath91 is : @xmath180 these functions are @xmath181-periodic and are defined on @xmath182 .
these functions diverge at @xmath183 but this is not an issue : the motion can diverge with respect to @xmath117 , but , according to the time @xmath96 and the integration ( see section [ time ] ) , the @xmath20-motion remains continuous for @xmath184 as @xmath1850;4k(k_w)[\longleftrightarrow \mathcal{\tau}\in\mathbb{r}\ ] ] let us assume the initial conditions provide @xmath164 , then : @xmath186 with @xmath187 the sign depends on @xmath166 : @xmath167 if @xmath166 is positive ( resp .
negative ) , @xmath91 increases ( resp . decreases ) and @xmath188 increases ( resp . decreases ) too .
then , @xmath169 is positive ( resp .
negative ) .
this solution corresponds to the solution @xmath189 given by @xcite .
these expressions are useful only for three real roots .
a last case remains : only one real positive root .
for any initial conditions , the particle will be escaping if @xmath59 has only one real root .
the solution for this case is more complex compared with the previous solutions .
as done by @xcite , we could apply some transformations and obtain a new expression , that would be a combination of @xmath140 and @xmath141 .
nevertheless , we propose a simpler way to determine the expression with the knowledge of all roots , even imaginaries .
we start from the equation [ wdiff ] with @xmath155 and @xmath156 not real : @xmath190 after the separation of variables , we obtain : @xmath191 now , we apply the procedure proposed in @xcite ( p. 597 ) in the case where we have only one real root .
first , we define : @xmath192 and also @xmath193 according to @xcite ( p. 597 ) , the left hand side corresponds to : @xmath194 with @xmath147 the elliptic function of first kind ( defined in the appendix [ appendixa ] ) and @xmath195 according to the equation [ equation60 ] : @xmath196 with @xmath197 the jacobi amplitude defined as @xmath198 thus @xmath199 finally , one can simplify the expression based on @xcite ( equation 22.6.18 ) : @xmath200 this function is @xmath201-periodic and is defined on @xmath202 .
this function diverges at @xmath203 but this does not impact our result : the motion can diverge with respect to @xmath117 , but , according to the time @xmath96 and the integration ( see section [ time ] ) , the @xmath20-motion remains continuous for @xmath184 as @xmath204 - 2k(k_{w_0});2k(k_{w_0})[\longleftrightarrow \mathcal{\tau}\in\mathbb{r}\ ] ] as usual , according to the initial conditions , we define @xmath205 as @xmath206 with @xmath207 if @xmath102 is positive ( resp .
negative ) then @xmath205 is negative ( resp . positive ) .
this solution corresponds to the solution @xmath151 given by @xcite but our derived solution is less complex to use ( with less time computing ) .
we gave analytical formulations for the different kinds of trajectories , expressed implicitly .
now , since we have all expressions of the trajectories as a function of @xmath117 , we can express the real time @xmath96 ( or @xmath22 ) : @xmath208 here , we use mathematica and maple to derive the primitives .
it is necessary to be very careful since these different programs can have different definitions of the elliptic functions for example . to avoid these issues , we remind at each use the definition employed .
the first part of the integral gives : @xmath209 \end{array } \label{tu}\ ] ] with @xmath210 , @xmath25 the incomplete elliptic function of the second kind ( see appendix , eq .
[ goodformule ] ) .
the second part of the integral is more complex because we have different expressions according to the number of roots and the initial conditions . in the case of three real roots , if @xmath211 then @xmath212 \end{array } \label{tw1}\ ] ] with @xmath213 .
if @xmath214 with three real roots then @xmath215\\\\ -\dfrac{\beta}{\lambda_a}\left[\dfrac{{\mathrm{cn}}(\beta(\mathcal{t}_0-\mathcal{t}_w),k_w)}{{\mathrm{sd}}(\beta(\mathcal{t}_0-\mathcal{t}_w),k_w ) } -\dfrac{{\mathrm{cn}}(-\beta\mathcal{t}_w , k_w)}{{\mathrm{sd}}(-\beta\mathcal{t}_w , k_w)}\right ] \end{array } \label{tw2}\ ] ] finally , in the case of only one real root , the time equation is given by : @xmath216\\\\ + \dfrac{\gamma}{\lambda_a}\left[\dfrac{{\mathrm{sn}}(2\gamma(\mathcal{t}_0-\mathcal{t}_{w_0}),k_{w_0}){\mathrm{dn}}(2\gamma(\mathcal{t}_0-\mathcal{t}_{w_0}),k_{w_0})}{1+{\mathrm{cn}}(2\gamma(\mathcal{t}_0-\mathcal{t}_{w_0}),k_{w_0 } ) } -\dfrac{{\mathrm{sn}}(-2\gamma\mathcal{t}_{w_0},k_{w_0}){\mathrm{dn}}(-2\gamma\mathcal{t}_{w_0},k_{w_0})}{1+{\mathrm{cn}}(-2\gamma\mathcal{t}_{w_0},k_{w_0})}\right ] \end{array } \label{tw3}\ ] ] with @xmath217 . for both last cases , @xmath218 belongs to @xmath2190;4k(k_w)[$ ] ( respectively @xmath219 - 2k(k_{w_0});2k(k_{w_0})[$ ] ) for the equation [ tw2 ] ( resp . for the equation [ tw3 ] )
nevertheless , when @xmath218 tends to @xmath201 ( resp .
@xmath220 ) , the integration diverges and @xmath96 too .
the @xmath20-motion occurs on an subset of @xmath221 with respect to @xmath117 but on @xmath221 entirely with respect to @xmath96 .
the transformation of @xmath96 into @xmath117 is bijective because the integrand @xmath222 is strictly positive .
to complete the description of the motion as a function of time , it is also necessary to solve the evolution of the angle @xmath10 , obeying to : @xmath223 as already done in the previous part , we separate into two integrals .
the first part still gives : @xmath224 with @xmath225 the incomplete elliptic integral of the third kind ( see appendix , eq .
[ goodformule ] ) . in the three real roots case for @xmath59 , if initially we have @xmath211 : @xmath226 or if initially @xmath214 : @xmath227 and for the last case with one real root : @xmath228 this last formula is only available for @xmath229 - 2k(k_{w_0});2k(k_{w_0})[$ ] , range where it is continuous .
with the solutions previously derived , we can know the exact motion of a bounded or unbounded particle as a function of the time such as given in figure [ sample ] .
it is clear that even a bounded trajectory the motion has no periodicity at all ( especially for the @xmath10 motion ) .
nevertheless , it could be interesting to focus on stable bounded orbits and search for periodic motions ( as in @xcite ) , and thus investigate in particular the circular stable orbits for spacecrafts @xcite or the possible positions for satellite particles produced by collisions in the exosphere @xcite .
thus , we dedicate this section to the conditions to obtain such orbits . for a specific set of initial conditions , it can be possible to obtain circular orbits .
this orbit occurs when , on the one hand , the attraction of the planet projected along the @xmath230-axis is equal to acceleration due to the radiation pressure : @xmath231 in dimensionless quantity , this will be expressed by : @xmath232 on the other hand , it is also necessary that the centrifugal force induced by the rotation around the @xmath230-axis is equal to the acceleration around the planet in the perpendicular plane to the @xmath230-axis .
thus , we obtain the secondary equality : @xmath233 in dimensionless quantity , this will be expressed by : @xmath234 combining these two equations , we obtain : @xmath235 we need to study the polynom @xmath236 with @xmath237 $ ] .
depending on the @xmath238 values , we have zero , one or two solutions for @xmath239 . indeed ,
@xmath240 so that according to the rolle theorem , there is @xmath2410;1[$ ] with @xmath242 . here
, @xmath1 is equal to 8/9 .
nevertheless , if @xmath243 is positive then @xmath239 does not have solutions and inversely , if @xmath243 is negative then we have two solutions .
this value is : @xmath244 a critical maximum value of @xmath245 thus exists to allow for circular orbits and is @xmath246{3}}\ ] ] above this value , we can not find any bounded trajectories : there is no any equilibrium point and thus no circular orbits ( stable or not ) .
for lower values of @xmath247 , we have two solutions for @xmath239 : one stable and one unstable as shown by @xcite and theirs plots of the equipotentials .
these two solutions correspond respectively to the stable point around which the equipotentials are closed and to the saddle point , which is the last limit where one can find closed equipotentials , and is the only point where two equipotentials can cross . as long as @xmath248 ,
these both specific points exist : they have the same @xmath90 .
physically , the potential @xmath57 has two extrema as plotted in figure [ potentiel ] .
when @xmath247 reaches @xmath249 , the local minimum goes to the right and the local maximum goes to the left at the same location @xmath250 . for higher @xmath247 values ,
@xmath56 have no extremum anymore : the potential is strictly decreasing and the particles are unbounded ( escaping ) .
thus , the bounded particles , satellite and ballistic particles , have @xmath248 .
the distinction between them thus depend on if they cross or not the exobase .
the critical values are given in @xmath251 coordinates ( @xmath252 is @xmath253 for us , in comparison with @xcite ) .
in dimensionless unities and in @xmath254 using the equations [ centrifuge ] then [ r ] , the critical orbit is : @xmath255 or in @xmath256 coordinates : @xmath257 the real positive roots of the polynomial [ polynom ] combined with the equality [ r ] give the positions of the circular orbits ( two coordinates are necessary ) allowed to spacecraft or particles under the influence of both gravity and radiation pressure .
the knowledge of the exact trajectories of particles or satellites under the influence of gravity and radiation pressure needs the calculation of the spatial coordinates , i.e. the @xmath258 motions , as well as the time evolution .
we summarize all needed equations in table [ summary ] . ' '' '' & & & + ' '' '' & & & + ' '' '' & & & & + ' '' '' & [ umotion](l2d ) & & & + ' '' '' & & [ wmotion1](l2d)&[wmotion2](l2d)&[wmotion3](s ) + ' '' '' & [ tu](l2d)&[tw1](l2d ) & [ tw2]*new*&[tw3]*new * + ' '' '' & [ phiu](l3d)&[phiw1](l3d ) & [ phiw2]*new*&[phiw3]*new * + ' '' '' & @xmath259&@xmath260&@xmath261&@xmath262 + the @xmath90-motion is provided by the equation [ umotion ] .
the @xmath91-motion is provided by the equations [ wmotion1 ] , [ wmotion2 ] or [ wmotion3 ] .
the @xmath10-motion is provided by @xmath263 $ ] .
the time equation is provided by @xmath264 $ ] .
all the expressions are functions of @xmath117 that is not the real time .
we thus have implicit expressions as a function of time .
the function @xmath265 is bijective but can not be inversed analytically , a numerical inversion is needed to derive the real time .
plane ( upper left panel ) , of the time as a function of @xmath117 ( upper right panel ) , of the motion in polar coordinates @xmath266 ( lower left panel ) and the @xmath267 coordinates as a function of the time ( red for @xmath90 , blue for @xmath91 , lower right panel ) . @xmath90 and @xmath91 do not show any periodicity because their periods are not commensurable ( i.e. the ratio is a rational number ) with @xmath117 and thus with the time @xmath96,title="fig : " ] plane ( upper left panel ) , of the time as a function of @xmath117 ( upper right panel ) , of the motion in polar coordinates @xmath266 ( lower left panel ) and the @xmath267 coordinates as a function of the time ( red for @xmath90 , blue for @xmath91 , lower right panel ) . @xmath90 and @xmath91 do not show any periodicity because their periods are not commensurable ( i.e. the ratio is a rational number ) with @xmath117 and thus with the time @xmath96,title="fig : " ] + plane ( upper left panel ) , of the time as a function of @xmath117 ( upper right panel ) , of the motion in polar coordinates @xmath266 ( lower left panel ) and the @xmath267 coordinates as a function of the time ( red for @xmath90 , blue for @xmath91 , lower right panel ) . @xmath90 and @xmath91 do not show any periodicity because their periods are not commensurable ( i.e. the ratio is a rational number ) with @xmath117 and thus with the time @xmath96,title="fig : " ] plane ( upper left panel ) , of the time as a function of @xmath117 ( upper right panel ) , of the motion in polar coordinates @xmath266 ( lower left panel ) and the @xmath267 coordinates as a function of the time ( red for @xmath90 , blue for @xmath91 , lower right panel ) . @xmath90 and @xmath91 do not show any periodicity because their periods are not commensurable ( i.e. the ratio is a rational number ) with @xmath117 and thus with the time @xmath96,title="fig : " ] besides , our 3d solutions can be easily applied to the 2d case . indeed , in the 2d case , @xmath268 and thus
, one of the roots for each polynomial @xmath58 and @xmath59 is null : it could be @xmath269 or @xmath60 for @xmath58 ( if @xmath270 , there is no possible motion ) and any roots of @xmath59 .
we precise that in this case the @xmath10-motion is not important because the motion is planar .
compared with @xcite , our formulations are first developed for the 3d case and can be used easily for the 2d case , whereas @xcite gave only the methodology to obtain the 3d solutions based on 2d ones but not the expressions , which apparently leads to complex expressions .
@xcite provided also the exact formulas for bounded and unbounded trajectories using the weierstrass functions but this formulation is also difficult particularly because of the need to use the inverse weierstrass function , not implemented in all computer softwares and the need to work with complex values ( e.g. the complex logarithm function ) . in this paper
, we solved the motion for the bounded and unbounded trajectories in the 3d case ; we provide the exact formulas for all cases , as well as the definitions used in [ appendixa ] .
we also highlight in table [ summary ] which solutions are simpler compared with @xcite , which are completely new and which were only provided in the 2d case by @xcite .
moreover , beyond the new exact solutions given in this paper , the derivation of our solutions based on jacobi elliptic functions allows a good computing time and accuracy .
@xcite compared three types of solutions for the stark effect : two exact ones , proposed by @xcite ( jacobi elliptic functions ) and @xcite ( weierstrass elliptic functions ) , and a numerical one by @xcite ( based on taylor series ) .
they compared the cpu time , the number of calls for each analytic elliptic function and the accuracy between @xcite and @xcite .
even if we do not agree with the number of evaluations of each jacobi elliptic function mentioned by @xcite ( e.g. to call @xmath140 and @xmath141 is similar to call @xmath197 then @xmath271 and @xmath272 : @xmath273 and @xmath274 ) , two arguments show our solutions are efficient in terms of cpu time and accuracy : first , the solutions expressed in terms of jacobi elliptic functions ( such as in this paper or by @xcite ) are more efficient than weierstrass elliptic functions ( used by @xcite ) ; second , several solutions given above are less complex to implement than those by @xcite , e.g. equation [ wmotion3 ] .
also , the analytical formulations are preferable for long duration motions .
we determined analytically the trajectories of the particles or spacecraft under the influence of both planetary gravity and stellar radiation pressure .
we thus provide for the first time the complete exact solutions of the well - known stark effect ( effect of a constant electric field on the atomic hydrogen s electron ) with jacobi elliptic functions , for both bounded and unbounded orbits .
these expressions may be implemented for modeling spacecraft or particles trajectories : instead of solving the equation of the motion , based on differential equations , with numerical methods such as the runge - kutta method where one cumulates errors along the time , it is here possible to obtain precise expressions of the motion with only periodic errors , due to the precision on the evaluation of the elliptic functions used . in particular
, we provide the analytical conditions for stable circular orbits .
moreover , we discuss about the possible issues inherent to the formalism used and the importance of being extremely careful with the routines implemented . the formalism used here
will allow us in a next paper to generalize the work by @xcite to derive the exact neutral densities and escape flux in planetary exospheres , under the influence of both gravity and stellar radiation pressure .
this is important for understanding the atmospheric structure and escape of planets in the inner solar system , as well as the atmospheric erosion during the early ages where the radiation pressure ( and uv flux ) of the sun was extreme .
this work was supported by the centre national dtudes spatiales ( cnes ) .
in this paper , we use the three incomplete elliptic integrals @xmath147 , @xmath25 and @xmath225 : @xmath275 these expressions are not exactly @xmath25 , @xmath147 and @xmath279 .
they agree with the previous formulas [ goodformule ] in the range @xmath280-\pi/2;\pi/2[$ ] .
@xcite did not precise which formulations they used . according to theirs formulas and results , they used the left - hand side of the equations [ badformule ] .
this may be a problem for bounded trajectories : for @xmath281 , the integrals [ badformule ] are not continuous contrary to [ goodformule ] .
depending on the computer software , the routines and the definitions used for these functions , the results can show some issues ( e.g. no continuous motion )
. 16 natexlab#1#1url # 1`#1`urlprefix[2][]#2 [ 2]#2 , , . . , .
, , , , , , . . ,
. , . . , . , ,
. . , . , , ,
. . , . , ,
. , . . , . ,
. . , , , ,
. edition .
, , , . . ,
. edition . , . . | the planetary exospheres are poorly known in their outer parts , since the neutral densities are low compared with the instruments detection capabilities .
the exospheric models are thus often the main source of information at such high altitudes
. we present a new way to take into account analytically the additional effect of the radiation pressure on planetary exospheres . in a series of papers , we present with an hamiltonian approach the effect of the radiation pressure on dynamical trajectories , density profiles and escaping thermal flux .
our work is a generalization of the study by @xcite . in this first paper
, we present the complete exact solutions of particles trajectories , which are not conics , under the influence of the solar radiation pressure .
this problem was recently partly solved by @xcite and completely by @xcite .
we give here the full set of solutions , including solutions not previously derived , as well as simpler formulations for previously known cases and comparisons with recent works .
the solutions given may also be applied to the classical stark problem @xcite : we thus provide here for the first time the complete set of solutions for this well - known effect in term of jacobi elliptic functions .
exosphere , radiation pressure , stark effect , trajectories |
the susy n=1 kdv model is related to the following l - operator : @xmath1 where @xmath2 , @xmath3 , @xmath4 are the chevalley generators of twisted affine lie superalgebra @xmath5 , @xmath6 is a superderivative , the variable @xmath7 lies on a cylinder of circumference @xmath8 , @xmath9 is a grassmann variable , @xmath10 is a bosonic superfield with the following poisson brackets : @xmath11 . making a gauge transformation of the l - operator
we obtain a new superfield @xmath12 = @xmath13 , where @xmath14 and @xmath15 generate the superconformal algebra under the poisson brackets : @xmath16 the susy n=1 kdv system has an infinite number of conservation laws and the first nontrivial one gives the susy n=1 kdv equation : @xmath17 the integrals of motion are generated by the logarithm of the supertrace of the corresponding monodromy matrix , which has the following form : @xmath18 @xmath19 \big).\ ] ] its quantum generalization can be represented in the quantum p - exponential form ( for the explanation of this notion see below and @xcite for details ) : @xmath20 vertex operators @xmath21 are defined in the following way @xmath22 .
the universal r - matrix with the lower borel subalgebra represented by @xmath23 is equal to @xmath24 . due to this fact @xmath25 satisfies the rtt - relation : @xmath26 where @xmath27 , @xmath28 mean that the corresponding object is considered in some representation of @xmath29 .
thus the supertraces of the monodromy matrix ( `` transfer matrices '' ) @xmath30 commute , providing the quantum integrability .
it is very useful to consider the evaluation representations of @xmath29 , @xmath31 , where now the symbol @xmath27 means integer and half - integer numbers . denoting @xmath32 as @xmath33
we find that @xmath34 commute : @xmath35=0.\nonumber $ ] the expansion of @xmath36 in @xmath37 ( the transfer matrix in the fundamental 3-dimensional representation ) is believed to give us as coefficients the local i m , the quantum counterparts of the mentioned i m of susy @xmath38 kdv .
using the super q - oscillator representations of the upper borel subalgebra of the quantum affine superalgebra @xmath39 we define the @xmath40 operators ( see @xcite for details ) .
the transfer - matrices in different evaluation representations can be expressed in such a way : @xmath41 where @xmath27 runs over integer and half - integer nonnegative numbers .
@xmath40 operators satisfy quantum super - wronskian relation : @xmath42 one should note , that we use only @xmath43-dimensional `` @xmath44-induced '' representations ( sometimes called atypical ) of @xmath45 .
it allows , however , to construct the fusion relations , see below . to construct the relations like baxter s ones we introduce additional `` quarter''-operators , constructed `` by hands '' from the @xmath46-operators : @xmath47 for odd integer @xmath48 .
the baxter s relations are : @xmath49 @xmath50 the fusion relations have the following form very similar to the @xmath51 case : @xmath52 but they are only `` fusion - like '' because the `` quarter''-operators do not seem to correspond to any representation of @xmath45 . the truncation of these relations for different values of @xmath53 , being the root of unity : @xmath54 , @xmath55 , @xmath56 has the following form : @xmath57 in the case when @xmath58 , where @xmath59 , @xmath60 there exists an additional number of truncations : @xmath61 @xmath62 these relations allow us to rewrite the fusion relation system in the thermodynamic bethe ansatz equations of @xmath63 type .
in this paper we studied algebraic relations arising from the integrable structure of cft provided by the susy @xmath64=1 kdv hierarchy .
the construction of the @xmath46-operator as a `` transfer''-matrix corresponding to the infinite - dimensional q - oscillator representation could be also applied to the lattice models .
the relations like baxter s and fusion ones will be also valid in the lattice case because they depend only on the decomposition properties of the representations . + in the following we also plan to study the quantization of @xmath65 susy kdv hierarchies , related with super - w conformal / topological integrable field theories .
i am very grateful to my supervisor prof .
p. kulish .
it is a pleasure to thank prof .
m. semenov - tian - shansky and f. smirnov for useful discussions and prof .
l. baulieu , b. pioline and lpthe , univ .
paris 6 for support and hospitality .
this work was supported by dynasty foundation and crdf grant rum1 - 2622-st-04 .
kulish , a.m. zeitlin , zapiski nauchn .
seminarov pomi , 291 ( 2002 ) 185 ( in russian ) , english translation in : journal of mathematical sciences ( springer / kluwer ) * 125 * ( 2005 ) 203 ; hep - th/0312158 .
kulish , a.m. zeitlin , phys . lett .
* b 581 * ( 2004 ) 125 ; hep - th/0312159 ; p.p .
kulish , a.m. zeitlin , theor .
* 142 * , 2005 , in press ; hep - th/0501018 .
kulish , a.m. zeitlin , phys . lett . * b 597 * ( 2004 ) 229 ; hep - th/0407154 .
p. p. kulish , a.m. zeitlin , nucl . phys .
* b * , 2005 , in press ; hep - th/0501019 . | the quantum susy n=1 hierarchy based on @xmath0 twisted affine superalgebra is considered . the construction of the corresponding baxter s q - operators and fusion relations is outlined .
the relation with the superconformal field theory is discussed .
one of the most famous integrable systems ( is ) is the korteweg - de vries hierarchy .
it is related with the superconformal field theory because its poisson brackets give the virasoro algebra and the involutive family of integrals of motion ( i m ) providing the integrability of the conformal field theory ( cft ) . since the late 1980s
the supersymmetric and fermionic extensions of the kdv system have been known ( see e.g. @xcite , @xcite , @xcite and references therein ) , which in turn are related with superconformal field theory ( scft ) . during the following years they were extensively studied on both the classical and the quantum level .
+ however , up to the present nobody has applied the most successful method in the theory of integrable systems , the so - called quantum inverse scattering method ( qism ) to these is . in this short paper
we demonstrate some algebraic tools giving possibility to study susy n=1 kdv via qism . |
it is well - recognized that the competition between the single - ion kondo effect and the ruderman - kittel - kasuya - yosida ( rkky ) interaction , the two inevitable forces in any kondo systems with more than one local impurity magnetic moment , plays a crucial role in correlated systems ranging from dilute magnetic alloys to heavy fermion compounds @xcite .
the indirect rkky interaction , namely the interimpurity interaction mediated by conduction electrons via a short - range ( on - site ) kondo coupling , oscillates and decays with the interimpurity distance @xmath2 and fermi momentum @xmath3 @xcite .
when the rkky interaction grows toward the strong antiferromagnetic limit , the quantum many - body ground state will evolve from a collective kondo singlet state@xcite into the interimpurity singlet state locked by the rkky interaction .
however , whether a distinct separation or a quantum critical point exists between the two singlet ground states has remained elusive .
this theoretical issue has a fundamental importance as it closely correlates with the critical divergence or scaling behavior of several physical quantities , or emergent energy scales in realistic materials@xcite .
it is remarkable that even for the simplest case with only two local impurities , i.e. , the two - impurity kondo model ( tikm)@xcite , the evidences for a quantum critical point separating the two distinct singlet states do not converge .
on the one hand , the early numerical renormalization group ( nrg ) studies@xcite revealed an unstable fixed point characterized by diverging impurity staggered susceptibility and coefficient of specific heat at a finite ratio of the rkky interaction to the kondo temperature ( @xmath4 ) . on the other hand ,
the subsequent well - controlled quantum monte carlo ( qmc ) studies did not find such a divergence , instead , a crossover behavior in the corresponding quantities at very low but still finite temperatures was observed @xcite . since then , although the effective field theory analysis suggested the occurrence of critical points in several variants of the tikm @xcite , the strong debate in the tikm studies , especially in numerical studies , has still remained until today @xcite .
almost all the previous investigations , except for the qmc studies @xcite , rely on a decomposition of the conduction electrons into odd / even channels with respect to the impurity center .
under such a decomposition , the original two - impurity problem is mapped effectively onto two - channel / two - impurity problem defined on the momentum or energy space , resulting in various impurity couplings which are energy - dependent in general .
the obtained two - channel / two - impurity model , capturing the low - energy properties of the original problem , provides a base for the nrg and effective field theory studies .
in particular , the early nrg calculations @xcite assume `` energy - independent '' coupling constants in odd / even channels .
the qmc simulations @xcite , on the other hand , suffer from finite temperatures .
other nrg@xcite or dmrg@xcite studies , where no evidence was found for a quantum critical point after taking into account the `` energy - dependent '' coupling strengths or starting from a real - space two - impurity lattice model , indeed show the importance of lattice geometry details .
more seriously , these numerical results seem to strongly contradict a general phase shift argument based on the local fermi liquid theory@xcite , which states that a phase transition must exist between the two stable fixed points if a tikm preserves particle - hole ( ph ) symmetry @xcite . as a matter of fact , there are two types of ph symmetries associated with a standard tikm , corresponding to the cases with the interimpurity distance @xmath2 being even or odd , or in the lattice case , the impurity mirror center being on site or bond , respectively .
it is the first type of ph symmetry , namely the distance @xmath2 being even or the mirror center being on site , that can guarantee a phase transition @xcite .
although the phase shift argument comes from the odd / even decomposition with a number of simplifications including the spherical plane wave and linear dispersion approximations , its validity should not depend on ignoring energy dependence of coupling constants and other lattice details . otherwise , if there is no any phase transition in this case , it would indicate a breakdown of the local fermi liquid picture of the single - impurity kondo problem . here
we would like to emphasize that such a phase transition could be either a first - order transition or a quantum critical point . in the case of only the second type of ph symmetry
preserved , there is no such a guarantee , in other words , either a phase transition or a crossover takes place between the two stable fixed points . in our point of view
, the most straightforward approach to challenge or confirm this argument and further clarify the discrepancies among various numerical studies is to directly solve the ground state of a standard tkim for sufficiently large systems without using decomposition .
in fact , affleck et al.@xcite already outlined three conditions for such a decisive numerical study : ( 1 ) the studied model preserves the first type of ph symmetry ; ( 2 ) a model parameter is varied to pass the interimpurity singlet to the kondo singlet ; ( 3 ) sufficiently low temperatures are accessible and other model parameters are fine - tuned . in this paper , we reexamine whether or not there is a quantum critical point in a standard tikm by applying a newly developed numerical method , i.e. , the natural orbitals renormalization group ( norg)@xcite . different from the conventional nrg , the norg keeps faithfully all the lattice geometry details and does not need a decomposition into odd / even channels and mapping onto the momentum or energy space .
in particular , the ground state wave function as well as the relevant low energy excitations in electronic bath with several impurities can be calculated accurately in a well - controllable manner for a very large system .
therefore , the norg is particularly pertinent for studying the zero temperature quantum phase transition in the two - impurity kondo problem . in the following we will present the numerical results and discuss the main observations .
a brief description of the norg method and an illustration of the phase shift argument and other relevant numerical results are given in the appendixes .
our studied tikm is a standard one - dimensional lattice model@xcite , as schematically shown in fig .
[ model ] , which is described by the following hamiltonian , @xmath5,\\ h_{{\rm{kondo}}}=j\sum\limits_{j = 1,2 } { { { \bf{s}}_j}\cdot { \bf{s}}\left ( { { { \bf{r}}_j}}\right)},\\ h_{{\rm{rkky}}}=k{{\bf{s}}_1 } \cdot { { \bf{s}}_2 } , \end{array}\ ] ] where @xmath6 describes the non - interacting conduction band with @xmath7 being the annihilation operator of a conduction electron located at the @xmath8-th site with spin component @xmath9 and @xmath10 being the nearest - neighbor hopping integral , @xmath11 is the short - range ( on - site ) kondo coupling between each impurity spin @xmath12 and the spin @xmath13 of a conduction electron passing by the @xmath14-th impurity site , and @xmath0 is the direct exchange interaction between the two impurity spins .
such a tikm can be realized in nanoscale devices where the observed kondo signature varies with tunable rkky interaction@xcite .
notice that in realistic materials the direct interaction @xmath0 , which is finite even at vanishing kondo coupling , can be mediated as a superexchange via other filled orbitals as in some layered @xmath15-electron compounds@xcite . as shown in fig .
[ model ] , the two impurities are located @xmath16 lattice spacings apart .
the model exhibits the lattice inversion symmetry with respect to the mirror center of the two impurities .
as illustrated by affleck et al.@xcite , at half filling of the conduction band the model exhibits the first ( or second ) type of ph symmetry when @xmath2 is even ( or odd ) . throughout the present study we fix @xmath17 and take half filling of the conduction band , while the periodic boundary condition is imposed and the length of the chain is denoted by @xmath18 .
when @xmath19 , the model becomes the one studied by fye et al .
using qmc method @xcite .
motivated by the suggestion of affleck et al.@xcite , we mainly consider the two representative cases , i.e. , the interimpurity distance being odd ( @xmath20 ) or even ( @xmath21 ) , respectively .
we note that the case with even @xmath2 was not thoroughly considered in the previous qmc calculations@xcite .
however , the required first type of ph symmetry is realized only in this situation .
we also notice that actually in most of the previous numerical studies @xcite only indirect rkky interaction is considered , corresponding to @xmath22 . in the present work ,
the eigenvalues and wave functions of the ground state and low - lying states of hamiltonian ( [ eq - hmlt ] ) were first directly solved by the norg method for sufficiently large systems .
the relevant physical quantities were then calculated . here
the proper physical quantities to characterize the ground state structure of the system and describe its response to external field are respectively the interimpurity spin - spin correlation function @xmath23 ( @xmath24 denotes ground state expectation value ) and impurity staggered susceptibility @xmath25 that is defined as @xmath26\left [ { s_1^z - s_2^z } \right ] } \right\rangle } $ ] with @xmath27 . to monitor the change of physical property of the system upon varying couplings @xmath0 and @xmath11 , we thus calculated @xmath28 and @xmath29 as functions of @xmath0 and @xmath11 .
the corresponding quantum phases were achieved by extrapolating @xmath28 and @xmath29 to the thermodynamic limit , namely @xmath30 , as described in appendix c for details . at @xmath19 and at @xmath31 and @xmath32
respectively for @xmath33 .
here @xmath34 , and @xmath35 respectively .
the endpoint at @xmath36 of each curve is determined by a quadratic polynomial fit of the leftmost four points of that curve calculated by norg . here
the norg results are consistent well with those of fye s qmc@xcite.,width=302 ] as a benchmark test , we carried out the calculations for a set of parameters with @xmath22 , @xmath37 , @xmath38 and @xmath32 , and with varying lattice sizes @xmath18 , respectively .
figure [ chi.r1k0 ] shows the calculated impurity staggered susceptibilities along with finite size extrapolations to the thermodynamic limit . as we see , the results in the thermodynamic limit fall within the error bars of those calculated by the early qmc @xcite .
this also confirms the early qmc results .
we remind that the case of @xmath39 was not studied by qmc calculations since this would introduce technical difficulties in qmc . by contrast
, there is no extra difficulty by using the norg approach , as shown below . and
interimpurity spin - spin correlation @xmath40 ( inset ) as functions of @xmath0 for @xmath41 and @xmath42 with different lengths @xmath18 .
it is similar to the case of @xmath41 and @xmath43 ( not shown ) .
for @xmath44 represented by solid lines @xcite , @xmath45 shows a divergent peak at @xmath46 which moves to the right and finally converges to @xmath47 as @xmath48@xcite . and
for @xmath49 represented by dotted lines , @xmath45 shows a peak ( with discontinuity ) at @xmath50 which moves to the left and finally converges to the same @xmath51 ( determined in the @xmath44 case ) as @xmath48 and meanwhile the peak height increases and finally diverges .
@xmath40 jumps where @xmath45 peaks .
the discontinuity of @xmath52 indicates a quantum phase transition , which is further characterized as a quantum critical point by the divergence of @xmath45 .
, width=302 ] we first study the cases of interimpurity distance @xmath53 and 4 with kondo coupling @xmath54 respectively , in which there is the first type of ph symmetry .
figure [ r2j1 ] shows the calculated impurity staggered susceptibility @xmath29 and interimpurity spin - spin correlation @xmath55 at zero temperature as functions of @xmath0 for various lattice sizes .
physically a very large negative @xmath0 , namely a large ferromagnetic coupling , tends to lock the two impurity spins into a triplet , characterized by @xmath55 being positive and close to the triplet value @xmath56 .
this has been shown in the inset of fig .
the resulting triplet is then screened by the conduction electrons through the two - stage kondo effect process@xcite .
in particular , we see that this tendency extends to @xmath22 .
this means that the _ indirect _ rkky interaction , which is due to the second or higher order corrections to the expectation value of @xmath55 between two impurity spins ( when @xmath22 ) , is _ ferromagnetic _ at half filling .
thus the ground state for @xmath22 is always in the kondo regime . in order to reach the border of the interimpurity singlet state regime , we tuned the strength of @xmath0 in a wider range from ferromagnetic to antiferromagnetic . versus kondo coupling @xmath11 : calculated critical point @xmath51 as a function of @xmath11 .
this defines a phase boundary separating the kondo singlet phase ( small @xmath0 ) from the interimpurity singlet phase ( large @xmath0 ) at zero temperature .
@xmath57 is quadratic at the small @xmath11 limit ( lower inset ) and linear at the large @xmath11 limit ( upper inset).,width=302 ] for a large positive @xmath0 , namely a large antiferromagnetic coupling , the ground state shall escape from the kondo screening phase , entering the interimpurity spin singlet phase .
this is clearly seen from the interimpurity spin - spin correlation @xmath58 ( the inset of fig .
[ r2j1 ] ) , which is negative and approaching the singlet value @xmath59 on the large positive @xmath0 side . in this case , there is the development of interimpurity singlet driven by the antiferromagnetic correlation . in the intermediate regime of @xmath60
, however , we find a rather sharp peak in the impurity staggered susceptibility . with increasing lattice size
, the peak goes diverging and the position of the peak converges to @xmath51 , indicating that the divergence of impurity staggered susceptibility is an intrinsic property free of the finite size effect .
furthermore , when @xmath0 is around the location of the impurity staggered susceptibility peak , the corresponding interimpurity spin - spin correlation @xmath55 exhibits a concomitant jump across the onset of antiferromagnetic correlation signed by @xmath61 beginning to be negative .
this feature indicates a quantum phase transition taking place at @xmath62 from the kondo singlet regime to the interimpurity singlet regime , which is further characterized as a quantum critical point by the divergence of the impurity staggered susceptibility . and @xmath42 with different lengths @xmath18 .
this gives rise to @xmath63 with @xmath64 .
@xmath65 and @xmath66 are determined by fitting data with formula @xmath67 @xcite.,width=302 ] the location of the critical point @xmath51 can be precisely determined by fitting the location of the divergent peak in the large @xmath18 limit , as illustrated in appendix c. figure [ kc ] shows the critical point @xmath51 as a function of kondo coupling @xmath11 , which defines a phase boundary separating the kondo singlet phase ( @xmath68 ) from the interimpurity singlet phase ( @xmath69 ) . as we see , the phase diagrams for interimpurity distance @xmath53 and @xmath70 are basically the same , while the function @xmath57 shows a quadratic behavior at the small @xmath11 limit and linear behavior at the large @xmath11 limit . to quantify the quantum criticality , we elaborately calculated the impurity staggered susceptibility @xmath71 as a function of @xmath0 in a range centered at the critical point @xmath51 with different system sizes , reported in fig .
[ criexp ] . using a standard numerical technique@xcite
, we can derive that @xmath71 diverges towards @xmath51 in power law rather than logarithmically in the thermodynamic limit , namely @xmath72 with @xmath73 .
hence the critical divergence is faster than the logarithmic one but still slower than @xmath74 , indicative of a continuous phase transition in computational studies@xcite .
this exponent @xmath75 has not been reported in literatures to our knowledge . and interimpurity spin - spin correlation @xmath40 ( inset ) as functions of @xmath0 for @xmath33 .
@xmath41 is fixed with different lengths @xmath18 .
for @xmath44 , @xmath45 looks diverging at some value of @xmath76 .
but the diverging tendency is suppressed and @xmath51 moves to @xmath77 as @xmath48 @xcite .
therefore , the @xmath45 curves for @xmath44 and @xmath49 both converge to the same one in the thermodynamic limit @xmath78 .
the case of @xmath79 is similar ( not shown ) .
, width=302 ] as a comparison , we likewise study the cases of interimpurity distance @xmath37 and 3 with kondo coupling @xmath54 respectively .
there is now the second type of ph symmetry , which however does not guarantee any phase transition . in fig .
[ r1j1 ] we report the calculated impurity staggered susceptibility @xmath29 and interimpurity spin - spin correlation @xmath61 as functions of the direct exchange interaction @xmath0 . similar to the cases of @xmath53 and 4 , for a large negative ( positive ) @xmath0 , the @xmath61 approaches the triplet ( singlet ) value , namely @xmath56 ( @xmath59 ) ( the inset of fig.[r1j1 ] ) , corresponding to the kondo ( interimpurity ) singlet regime .
however , there is no any jump or discontinuity for @xmath80 upon @xmath0 varying .
actually , as shown in the inset of fig .
[ r1j1 ] , the change of @xmath80 from the kondo singlet regime to the interimpurity singlet regime is very smooth across the middle value @xmath81 between the singlet and triplet values , indicating co - existence of the kondo and interimpurity singlets in this intermediate regime , in other words , a crossover taking place .
this can be further confirmed by examining the behavior of impurity staggered susceptibility @xmath29 as a function of @xmath0 as follows .
in addition , the @xmath80 , being negative at @xmath22 , shows that the _ indirect _ rkky interaction is antiferromagnetic at @xmath22 , contrary to the cases of @xmath53 and 4 . from fig .
[ r1j1 ] we see that the @xmath82 curves for @xmath49 converge to a smooth and non - divergent one as @xmath48 @xcite .
in contrast , each @xmath82 curve for finite @xmath44 seemingly has a divergent peak .
however , further calculations @xcite show that the value of @xmath0 at which @xmath83 diverges for @xmath44 moves to the negative infinity as @xmath48 .
it turns out that eventually the @xmath83 curves for both @xmath44 and @xmath49 converge to the same smooth and non - divergent one in the thermodynamic limit .
this indicates the absence of a quantum critical point .
this conclusion is in agreement with the previous qmc calculations for the case of @xmath33 ( or @xmath79 ) .
the studied tikm offers a new ingredient , namely the interimpurity exchange @xmath0 , which competes with the single - ion kondo effect of individual impurities . at zero temperature , two different regimes of the ground state , i.e. , the kondo singlet and the inter - impurity singlet , are respectively realized when @xmath0 is varied from the large ferromagnetic to the large antiferromagnetic .
crossover between these two regimes is expected to be a generic feature since the relevant degrees of freedom due to quantum impurities are finite .
thus it comes as a surprise when an unstable non - fermi liquid fixed point or a quantum critical point separating the two regimes was firstly evidenced by the nrg studies@xcite .
the existence of such a critical point in the tikm has been then debated by various approaches even in the presence of ph symmetry@xcite . using the norg calculations we have demonstrated that there is indeed a quantum critical point for a tikm with the first type of ph symmetry .
the criticality is characterized by the power - law divergence of the impurity staggered susceptibility @xmath29 , concomitantly accompanied by a jump of the interimpurity spin - spin correlation @xmath84 , and thus belongs to a genuine impurity quantum phase transition where the impurity contribution to the ground state energy becomes singular @xcite . moreover , as presented in above sections , we have also shown that to resolve this longstanding issue depends on not only the required ph symmetry but also the tunable direct rkky interaction . meanwhile ,
the position of critical point is closely related to the lattice details .
the newly developed norg method , which keeps lattice details and is free of decomposition and finite temperatures , makes examination on all these aspects possible . here
we would like to remind that in the previous nrg studies on tikm ( or the corresponding two - impurity anderson model ) the occurrence of a critical point requires vanishing even / odd channel asymmetry@xcite or energy - independence of coupling constants @xcite .
however , the even / odd channel asymmetry or energy - dependence of coupling constants is an inevitable consequence of decomposition when mapping a lattice tikm onto a continuous model in momentum or energy space . on the other hand , the phase shift argument @xcite for
a quantum phase transition in a tikm does not reply on even / odd channel asymmetry or energy - independence of coupling constants .
therefore , our norg study on a standard lattice tikm directly confirms the long - cherished prediction of the link between the first type of ph symmetry and the two - impurity kondo critical point @xcite .
the present study clearly shows that the previous numerical studies actually do not contradict each other for their own sake because the discrepancies are mainly due to the fact that they adopted different parameter regions or approximations .
in particular , the energy - dependence of the coupling constants appeared in these studies should not be an obstacle for studying the critical point provided that the first type of ph symmetry and antiferromagnetic interimpurity interaction are realized .
as the required symmetry occurs at half filling in a lattice model with reflection symmetry about a site ( or interimpurity distance is even ) , the two - impurity kondo critical point should be observed experimentally in realistic materials with tunable interimpurity exchange interaction .
finally , the norg method developed here can be also used to solve the multi - impurity models in two or three dimensions@xcite , or used as impurity solvers in the dynamical mean - field theory .
therefore we expect that the norg method can be used to study other challenging problems in correlated electron systems .
this work is supported by national natural science foundation of china ( grant nos .
91121008 and 11190024 ) and national program for basic research of most of china ( grant no .
2011cba00112 ) . j.d .
was also supported in part by the qianjiang scholarship of zhejiang province at hangzhou normal university ( under grant no .
2012qdl037 ) .
computational resources were provided by the physical laboratory of high performance computing in ruc and national supercomputer center in guangzhou with tianhe-2 supercomputer .
for a quantum impurity model , by generalizing quantum renormalization group approach into natural orbitals space through iterative orbital rotations , we can realize a numerical many - body approach , namely the norg@xcite , with polynomial , no longer exponential , computational complexity ( @xmath85 ) in the number of electron bath sites ( @xmath86 ) .
it turns out that dozens , or even hundreds , of bath sites can be dealt with in practice .
in addition , the norg works in a hilbert space constructed from a set of natural orbitals , thus it can work on a quantum impurity model with any lattice topological structure . in the present work ,
we have improved the efficiency of the norg method .
the computational complexity is further reduced from @xmath85 to @xmath87 , in which the impurity orbitals are no longer involved into orbital rotations , and hence the transformed hamiltonians become simpler .
this enables us to calculate the eigenvalues and wave functions of ground states and low - lying excitations for larger system sizes .
the largest system size we have reached is @xmath88 ( see appendix c )
. errors of the norg are controllable . in this paper
the relative errors of all data calculated by the norg are less than @xmath89 .
we reiterate the phase shift argument for a phase transition between the collective kondo screening singlet state and the rkky interaction - locked interimpurity singlet state in the presence of the first type of particle - hole ( ph ) symmetry@xcite . in the field theory treatment of the two - impurity kondo model ( tikm ) , a decomposition of electron field into even / odd components ( @xmath90 ) is introduced ( the spin degree of freedom is implied ) , both components are dependent on the quasi - particle energy @xmath91 measured from the fermi energy @xmath92 which is set to be zero .
the obtained effective hamiltonian , @xmath93 , takes a general form as follows , @xmath94 where @xmath95 ( @xmath96 ) are energy - dependent coupling constants reflecting the lattice details .
the first type of ph transformation of the fields implies @xmath97 while the second type of ph symmetry implies @xmath98 in the fermi liquid picture , the zero temperature fixed point can be characterized by the phase shifts ( @xmath99 ) for both components at the fermi energy @xmath100 .
this amounts to the following relations between the incoming and outgoing operators @xmath101 hence @xmath102 in the first type of ph symmetry , and @xmath103 in the second type of ph symmetry . in the limit of @xmath104 ,
both channels are in the kondo screening phase , indicating @xmath105 . while , in the limit of @xmath106 , the kondo effect is completely suppressed by interimpurity correlation , leading to @xmath107 .
therefore , by varying @xmath108 , a phase transition must take place at certain value of @xmath51 when the first type of ph is maintained , as sketched in fig .
[ shift](b ) .
by contrast , no phase transition is guaranteed for the second type of ph symmetry ( fig . [ shift](a ) ) here we would like to remind that for a lattice model with inversion symmetry and specified kondo coupling at half filling , the first or second type of ph symmetry will be preserved when the interimpurity separation @xmath2 is even or odd , respectively . in general , a long - range hybridization or kondo coupling can be introduced in order to maintain the required ph symmetry .
the observation described in the main text indicates that a tikm with the simplest on - site kondo coupling which preserves the required ph symmetry is enough to support an emergent quantum critical point . in the even ( odd ) channel of the conduction band at the fermi energy.,width=302 ]
for @xmath33 , @xmath31 , and @xmath19 at zero temperature .
the endpoint at @xmath36 of each curve is determined by a quadratic polynomial fit of the leftmost four points of that curve calculated by the norg .
our norg result is consistent well with that of fye s qmc@xcite.,width=302 ] _ impurity spin - spin correlation .
_ another benchmark test for our norg method is the impurity spin - spin correlation , @xmath80 , in the ground state for @xmath33 , where the qmc result is available .
we choose @xmath31 and @xmath19 , the same parameters as given in fye s qmc study@xcite .
the calculated result is shown in fig .
[ s1s2.r1j0.728k0 ] .
the system size varies as @xmath109 .
the endpoint at @xmath36 of each curve is determined by a quadratic polynomial fit of the leftmost four points of that curve calculated by the norg .
our norg result is consistent well with that of fye s qmc@xcite .
the relative difference between the two endpoint values is just as small as @xmath110 , which highlights the high accuracy of the norg method . _
@xmath113 when @xmath53 and @xmath114 .
_ to compare with the early nrg result , we calculate the ratio @xmath113 , where @xmath115 is the single - ion kondo temperature defined as@xcite @xmath116 where @xmath117 ( @xmath118 here ) is the conduction electron bandwidth and @xmath119 ( @xmath120 here ) the conduction electron density of states at the fermi level .
@xmath113 is then plotted in fig .
[ kctk ] . while the generic behavior is similar for @xmath53 and @xmath70 , quantitative difference between them appears when @xmath121 .
it is seen that in a wide range of @xmath11 , @xmath112 is close to the value predicted by the nrg studies , @xmath122.@xcite at which @xmath45 diverges for @xmath79 and @xmath44 .
dots denote the results calculated by the norg .
lines are determined by fitting the dots with the formula @xmath123 , where @xmath124 , @xmath125 , and @xmath126 are fitting parameters .
the @xmath51 behavior for @xmath33 is similar ( not shown ) .
this figure clearly shows that @xmath51 decreases to @xmath77 as @xmath48.,width=302 ] at which @xmath45 diverges for @xmath42 , @xmath41 , and @xmath44 .
the dots are calculated by norg .
the line is determined by fitting the dots with formula @xmath127 , where @xmath51 , @xmath124 , and @xmath128 are fitting parameters .
this perfect fitting shows that @xmath67 with @xmath66.,width=302 ] _ determination of @xmath51 when @xmath2 is even_. for even r the tikm has a quantum critical point at @xmath76 where the impurity staggered susceptibility @xmath45 diverges . for finite system size @xmath44
, @xmath45 diverges at @xmath129 .
we extract these @xmath130 from @xmath45 curves calculated by norg for different values of @xmath18 and extrapolate it to @xmath131 by fitting . to present the fitting details we show the typical case of @xmath41 in fig .
the perfect fitting shows that @xmath67 with @xmath66 . for a series of values of @xmath11
, we show the result in fig .
[ kcr4j1 ] . while @xmath11 spans across three orders of magnitude , @xmath51 across seven orders of magnitude .
this shows the robustness of the norg method . _
suppression of susceptibility peak when @xmath2 is odd . _
the impurity staggered susceptibility @xmath29 shows a sharp peak in the case of @xmath132 ( or @xmath37 ) when @xmath133 .
this feature seems different to what happens when @xmath134 where no sharp peak is detected . in order to clarify whether this is an intrinsic property of the ground state in the thermodynamic limit @xmath135 , we plot the peak position @xmath51 by varying the lattice size @xmath18 for different values of @xmath11 ( fig .
[ kc.r3 ] ) .
the result suggests that in the limit @xmath136 , the peak position @xmath51 of @xmath45 for @xmath44 moves to negative infinity and @xmath45 s for both @xmath44 and @xmath49 converge to the same one with a non - divergent behavior in the thermodynamic limit .
therefore , no phase transition takes place in the case of @xmath132 ( or @xmath37 ) . o. sakai and y. shimizu , excitation spectra of the two impurity anderson model .
i. critical transition in the two magnetic impurity problem and the roles of the parity splitting , _ j. phys
jpn . _ * 61 * , 2333 ( 1992 ) . here
we follow nothing but the convention to choose a one - dimensional lattice version of the tikm system as a representative mode system to study@xcite . as presented in ref . , the effectiveness of the norg is essentially irrespective of a model s topological structure .
in other words , the norg can handle one - dimensional , two - dimensional , and three - dimensional systems with any geometrical structures in the same efficient way .
higher dimensional systems will be studied subsequently .
h. pruser , p.e .
dargel , m. bouhassoune , r.g .
ulbrich , t .
pruschke , s. lounis , and m. wenderoth , inteplay between kondo effect and ruderman - kittle - kasuya - yosida interaction , _ nat .
commun . _ * 5 * , 5417 ( 2014 ) .
j. dai , j .- x .
zhu , and q. si , @xmath15-spin physics of rare - earth iron pnictides : influence of @xmath138-electron antiferromagnetic order on the heavy - fermion phase diagram , _ phys .
b _ * 80 * , 020505 ( 2009 ) . for one dimensional tight binding model @xmath6 in eq .
( [ eq - hmlt ] ) with periodic boundary condition and with @xmath139 sites ( @xmath140 being an integer ) , its single - particle energy spectrum consists of a non - degenerate lowest energy level , a non - degenerate highest energy level , and in - between @xmath141 doubly degenerate energy levels . thus , at half filling the system is of open shell ( namely metallic - like behavior ) if @xmath140 is even , while it is of closed shell ( namely semiconducting - like behavior ) if @xmath140 is odd .
it turns out that the model described by eq .
( [ eq - hmlt ] ) respectively with @xmath142 and @xmath143 sites shows a much different behavior when @xmath140 is small ( see fig .
nevertheless , the present calculations show that when @xmath144 in both cases the model realizes the same thermodynamic limit .
quantum impurity models involve a finite number of local impurity degrees of freedom and infinite number of bulk degrees of freedom .
the coupling between them permits sharp transitions at @xmath145 and the impurity contribution to the free energy is singular at the critical point @xcite .
it is also proposed that the schmidt gap of the entanglement spectra of impurity systems can serve as an alternative order parameter as recently manifested in a dimerization spin chain with two impurity spins@xcite . in our present study , this feature is clearly manifested in the jump of the interimpurity spin - spin correlation .
we can envision that by increasing the number of impurities to infinity the jump could be replaced by a continuous transition as in the kondo lattice case with a local quantum critical point.@xcite | the problem of two magnetic impurities in a normal metal exposes the two opposite tendencies in the formation of a singlet ground state , driven respectively by the single - ion kondo effect with conduction electrons to screen impurity spins or the ruderman - kittel - kasuya - yosida interaction between the two impurities to directly form impurity spin singlet .
however , whether the competition between these two tendencies can lead to a quantum critical point has been debated over more than two decades . here , we study this problem by applying the newly proposed natural orbitals renormalization group method to a lattice version of the two - impurity kondo model with a direct exchange @xmath0 between the two impurity spins .
the method allows for unbiased accessing the ground state wave functions and low - lying excitations for sufficiently large system sizes .
we demonstrate the existence of a quantum critical point , characterized by the power - law divergence of impurity staggered susceptibility with critical exponent @xmath1 , on the antiferromagnetic side of @xmath0 when the interimpurity distance @xmath2 is even lattice spacing , while a crossover behavior is recovered when @xmath2 is odd lattice spacing .
these results have ultimately resolved the long - standing discrepancy between the numerical renormalization group and quantum monte carlo studies , confirming a link of this two - impurity kondo critical point to a hidden particle - hole symmetry predicted by the local fermi liquid theory . |
in the past decade , carbon nanotubes ( cnts ) have been studied extensively because of their unconventional properties from both the fundamental research and application point of view @xcite . for the application to nanoscale electronic devices ,
researchers have fabricated various forms of cnts to engineer their physical properties .
@xmath2-bonded carbon nanostructures have been found to coalesce by electron beam irradiation at high temperature and new morphologies such as x- and t - shaped junctions have been produced @xcite .
these developments offer interesting opportunities to study phase - coherent transport in novel geometries . thus far ,
phase - coherent transport has been most actively studied in the semiconductor heterostructures by employing the aharonov - bohm ( ab ) interferometer @xcite with an embedded quantum dot in one arm .
the measured phase in the open geometry ( as in double - slit experiments ) is featured with a smooth increase at the resonant tunneling and an abrupt jump by 180@xmath3 at transmission zeroes . in the closed or unitary geometry
, the conductance generally exhibits the fano line shape @xcite .
the fano effect usually arises from the coherent interference between a narrow localized level ( quantum dot ) and a continuum energy spectrum ( the arm without a quantum dot ) @xcite .
the fano effect has also been identified in the direct transport through a single quantum dot @xcite .
the cnt is excellent for observing phase coherence phenomena and there are some experimental signatures of the fano effect in the cnts @xcite , though the detailed mechanism is not fully understood . in this paper , we report transport properties of a multiply connected carbon nanotube ( mccn ) structure as shown in fig .
1 , where a single tube is branched off into two smaller arms and then they merge into one @xcite .
we have studied conductance variation systematically by changing the length of the arms . both @xmath1-bonding and @xmath4(@xmath1 anti)-bonding electron transport channels show resonant tunneling through discrete energy levels(dels ) in the finite arms .
the width of the resonant tunneling peaks in the @xmath1 channel is broad and the transmission probability is fairly uniform as a function of energy . the @xmath4 channel , on the other hand , has more interesting structure of either broad or narrow resonant tunneling as will be shown later .
coherent interference between a very broad level ( extending over to nearby levels ) and its narrow neighboring level is evident in the asymmetric fano - type line shapes of the @xmath4 transport channel and the corresponding transmission probability is featured with both zero and unity . in particular , we find a nearly perfect suppression of the @xmath4 transmission in a certain finite range of energy .
this mechanism can be regarded as electron filtering of a particular wave function character ( @xmath1-pass filter ) , and such a nonequilibrium distribution may be maintained over the relaxation length of the electron phase beyond the arm region .
our model system , the mccn , comprises two leads of semi - infinite metallic ( 10,10 ) cnts and two arms of finite ( 5,5 ) tubes ( effectively a resistive region ) of the same length in between .
this structure may be thought as two y - junctions faced to each other , possibly produced experimentally by electron irradiation on nanotubes .
six heptagons are contained in each of two junction regions where the ( 10,10 ) tube and two ( 5,5 ) tubes are joined .
the length of the arm region is represented by the number ( @xmath5 ) of periodic units of the armchair cnt as shown in fig .
1 . this structure possesses the mirror symmetry with respect to three planes .
two symmetry planes @xmath6 and @xmath7 , which are , respectively , perpendicular to and containing the ( 10,10 ) tube axis , are indicated in fig .
1 . the third one is the plane of the figure containing the tube axis ( not indicated in fig .
the electronic structure is described by the single @xmath1-electron tight - binding hamiltonian @xmath8 where @xmath9 denote the nearest neighbor pairs , the hopping integral @xmath10 ev @xcite and the on - site energy is set to zero .
it is well known that the @xmath1 and @xmath4 bands are crossing at the fermi level ( @xmath11 ) in the armchair cnts as shown in fig .
a @xmath1-bonding state consists of @xmath12 orbitals having the same phase all across the circumference of the tube . in a @xmath1-antibonding ( @xmath4 ) state , on the other hand
, the constituent @xmath12 orbitals have alternating signs ( 180@xmath13 out of phase ) along the circumference of the tube . @xmath1 and
@xmath4 states are usually admixed when defects are introduced into the perfectly periodic nanotubes .
however , @xmath1 and @xmath4 transport channels remain unmixed in our mccn because of the @xmath14 symmetry of the structure . only if @xmath1 and @xmath4 states are unmixed , the effect of interference among different levels becomes pronounced and dramatic ( e.g. , complete suppression of transmission of a level ) as described below . furthermore , because two arms are of the same length , we can focus on other effects than the traditional ab interference due to the path length difference . in our scattering - state approach ,
we solve the schrdinger equation for the whole system by matching the solutions of the tight - binding hamiltonian in the arm region with that of the lead region at each interface .
for given incoming electron waves from the left lead as initial conditions , we obtain the transmitted electrons emerging on the right lead .
conductance is obtained from the landauer - bttiker formula , @xmath15 , where @xmath16 is the conductance quantum ( @xmath17 ) and @xmath18 is the transmission matrix @xcite .
we do not include here extra contact resistance that may arise between external metallic leads and the mccn .
the conductance as a function of the incident electron energy is displayed in fig . 3 for the cases of @xmath19 5 , 9 and 12 .
to examine the structure of the conductance in detail , total conductance is decomposed into two nonmixing contributions of @xmath1 and @xmath4 channels , @xmath20 , where @xmath21 and @xmath22 are the transmission probabilities of the two .
we find the following features common to various @xmath5 s .
( i ) @xmath21 has peaks of magnitude one always and varies very slowly as a function of energy .
( ii ) @xmath22 has both broad and narrow peaks of magnitude one always , and the line shape is highly asymmetric , especially for narrow peaks .
( iii ) @xmath21 has no zeroes within the interested energy window while @xmath22 is featured with zeroes near the narrow asymmetric peaks .
since the wave function of a @xmath1 state does not have phase variation in the circumferential direction ( i.e. , the same sign over the cross - section of the tube ) irrespective of @xmath23 in ( @xmath23,@xmath23 ) tubes , the @xmath1 wave function in the lead part ( ( 10,10 ) tube ) has almost perfect match with that in the arms ( two ( 5,5 ) tubes ) , and the transmission is close to unity in a wide range of energy .
in contrast , the conduction of @xmath4 electrons is more complicated because the wave function in the lead part can not match that of the two - arms region due to the phase variation along the circumferential direction as mentioned before .
the conducting behavior may be explained by a close examination of the resonant tunneling through the dels which are accommodated in the arms of ( 5,5 ) tubes .
each junction area acts as an effective scattering center for @xmath4 electrons .
thus the mccn structure may be regarded as a double - wall quantum well and we expect the energy quantization by confinement to a finite region of the arms . with increasing length @xmath5 of ( 5,5 ) tubes ,
the spacing between peaks becomes smaller .
peaks for the @xmath4 states in fig .
3 show this trend unambiguously .
( @xmath1 electrons also experience resonant tunneling via dels as exemplified in the series of peaks of magnitude unity
. however , their line width is so broad that the variation in the conductance is small and uninteresting . )
the spectrum of dels can be roughly estimated from the linear dispersion @xcite near @xmath11 for the ideal armchair nanotube , @xmath24 (= 2.46 ) is the lattice constant and @xmath25 ( @xmath26 ) is the fermi wave vector .
approximate values of the spectrum of dels are found from the quantization condition @xmath27 , where @xmath28 is the effective length of the arm and @xmath29 is positive integer .
for instance , if we took @xmath30 for @xmath31 , the dels near @xmath11 would be @xmath320.93 , @xmath320.37 , 0.19 , and 0.74 ev . a better fitting to the exact numerical calculation in fig .
4(a ) could be obtained if we assumed a longer effective length @xmath33 , in which case we would have @xmath320.82 , @xmath320.32 , 0.18 , and 0.68 ev .
this result is reasonable since the electrons localized in the central ring region are somewhat spread out beyond the arms as seen , for instance , in the last picture of fig .
this simple model based on the linear dispersion , of course , has a limitation of obtaining , incorrectly , the same energy separations only .
coupling to the left and right leads results in the finite lifetime ( line broadening ) of the dels . to understand the variation of the line width from peak to peak , we further investigate the structure of the wave functions at the resonant peaks .
besides the linear @xmath1 and @xmath4 bands crossing at @xmath11 , there exist a conduction band at 0.8 ev and above , and a valence band at @xmath320.8 ev and below in the ( 10,10 ) tube . because of the presence of the heptagonal carbon rings at the junction , donorlike states tend to be produced according to well - known hckel s ( @xmath34 ) rule for stability @xcite .
the cycloheptatrienyl cation is a molecular example of the heptagonal ring following hckel s rule for @xmath35 .
the localized impurity ( donor ) state is most pronounced if the energy level is close to ( and below ) aforementioned conduction band minimum at 0.8 ev .
such localized donorlike states have in fact been found in other defects ( nitrogen impurity and stone - wales defect ) of the nanotube as well @xcite .
the peak c in fig .
4 ( or sharp peaks closest to 0.8 ev in fig .
3 ) corresponds to this state and has a narrow line shape owing to the relatively well - localized ( long lifetime ) character of such an origin .
the dels further away from the conduction band minimum are less localized and thus broader .
we also compare the wave functions around the peaks and the nodes of the conductance curve .
the wave functions at peaks of @xmath22 have large amplitude in two arms so that incident electrons are in resonance with the states in the region .
in contrast , the wave functions of the @xmath0 band at the u - shaped valley ( e.g. , at 0.46 ev in fig .
4(b ) ) are depleted inside two arms , which means an effective disconnection of two leads and the suppression of the electron conduction .
unlike in the @xmath1 electron case , the phase variation across the cross - section of the tube is very rapid in the @xmath4 state and it is very difficult to match the wave function at the junction boundary for a general value of energy , hence substantial reduction in conductance .
so far , we have accounted for the interesting phenomenon of line width variations from level to level .
such a variation leads to even more profound consequences on the structure of the transmission probability as we shall describe below . in the 1d mesoscopic systems such as the double - barrier quantum wells , if all the dels are coupled to the continuum of lead states with almost the same coupling strength and the relative signs of couplings to two leads are alternating from level to level
, the transmission zero does not occur @xcite .
however , when one energy level is coupled by far more strongly to the leads and acts like a broad band interfering with other narrow levels , the transmission zero may take place @xcite .
our system falls in the latter case . as the energy of the dels
is increased , the number of nodes along the two arms increases one by one and in turn the relative signs of the wave functions at two junctions are oscillating for dels .
the dels in the arm region are coupled to the left and right ( 10,10 ) tube leads with the same coupling strength but with variable signs .
the signs of the coupling constants at both junctions are determined by the parity of the wave functions with respect to the reflection symmetry plane , @xmath6 . for an even - parity state ,
the signs of coupling constants at two junctions are the same , but they are opposite for an odd - parity state .
this alternating parity under @xmath6 symmetry operation is clearly visible in fig .
4(b ) for dels at @xmath320.32 ( even ) , 0.18 ( odd ) , and 0.59 ev ( even ) . in the presence of the @xmath6 symmetry ( plus the time - reversal symmetry which obviously exists in the absence of applied magnetic fields )
, it is well known that the perfect transmission ( @xmath36 ) is allowed to occur at all resonant peaks in the system @xcite .
on the other hand , the transmission zero is generally not allowed in the energy range between one level and its nearest neighbor level interfering each other @xcite .
it is only when a broad level extends beyond the nearest neighbor narrow level that a transmission zero occurs at the energy between its nearest and _ next nearest neighbor _ narrow level . to confirm this argument
, we fit the curve of @xmath22 for @xmath37 with the multilevel generalization of the green s function formula @xcite .
the expression for @xmath38 contains , as parameters , dels ( @xmath39 ) and coupling constants to right(@xmath40 ) and left leads(@xmath41 ) @xcite . @xmath42
@xmath43^{-1},\end{aligned}\ ] ] where @xmath44 is the coupling matrix between the resistive and lead region on the left or right , @xmath45 is the retarded or advanced green s function , @xmath46 , @xmath47 and @xmath48 is the density of states in the lead on the left or right ( @xmath49 in the present case ) .
we use seven dels and their diagonal components of @xmath50 .
values of dels ( @xmath51 ) in ev are as follows : @xmath321.330(0.139 ) , @xmath320.859(0.194 ) , @xmath320.300(0.132 ) , 0.178(0.025 ) , 0.584(0.013 ) , 1.100(0.183 ) , and 1.700(0.250 ) .
the conductance curve in much wider energy range than the one presented in fig .
4 has been fitted for more complete results . as mentioned above , relative signs of the wave function at two junctions alternate from one level to the next so that we set @xmath52 , where @xmath53 is 1 ( @xmath321 ) for the even ( odd ) parity state with respect to @xmath6 .
@xmath50 represents the broadening of a particular level and results in the finite energy width of the conductance peak . in fig .
4 , a , b , and c are the broadened peaks associated with respective dels . at a given energy , contributions from tails of
two or more broadened peaks overlap and give rise to an asymmetric line shape . for instance , between a and b ( i.e. , the right - hand side of a and left - hand side of b ) , conduction paths interfere constructively and no transmission zero occurs . however , on the right - hand side of b , a and b interfere destructively ( since the energy is greater than both a and b ) and produce a transmission zero . since peak c is very narrow , its constructive interference on b ( in the energy range between b and c ) is much less than the destructive interference from a and unable to change the result .
in fact , there are two transmission zeroes between b and c , one nearer to b which is obtained as above and the other nearer to c which is obtained by the destructive interference between c and the next broad peak above c ( not shown ) . here
we emphasize that the transmission zeroes result from the interference between the abstract conduction paths through the energy levels in the hilbert space .
the curve fitting with eq.(3 ) is practically indistinguishable from the @xmath4-channel conductance of our mccn , indicating that the simple model with the multilevel transmission formula is valid for explaining the fano resonance structures and the transmission zeroes .
we also observe that , in a certain finite energy range , the conductance of @xmath4 electrons is almost zero while that of @xmath1 electrons is not reduced appreciably .
for instance , the transmitted electrons of @xmath4 character for @xmath54 in fig .
3(a ) constitute less than 8% of the total transmitted electrons in the energy range between 0 and 0.5 ev .
this may be regarded as orbital filtering of a particular wave function character ( @xmath1-pass filter ) .
this result can be used for a further fundamental research on the orbital character in nanostructures or an application to current switching in nanotubes .
we have carried out tight - binding calculations of conductance of multiply connected metallic cnts with the mirror symmetry .
the @xmath1 channel is well - conducting and has very broad resonance peaks .
the @xmath4 channel exhibits fano resonance structures .
the fano effect originates from the interference between broad and narrow electronic levels in the arm region .
it is shown that the present system transmits predominantly @xmath1 over @xmath4 electrons in a certain energy range .
this work was supported by the cnnc of sungkyunkwan university , the most through the nstp ( grant no .
m1 - 0213 - 04 - 0001 ) , the grant no . 1999 - 2 - 114 - 005 - 5 from the kosef , the korea research foundation grant ( krf-2003 - 042-c00038 ) , the samsung sdi - snu display innovation program , and the supercomputing application support program of the kisti .
jinhee kim , jae - ryoung kim , jeong - o lee , jong wan park , hye mi so , nam kim , kicheon kang , kyung - hwa yoo , and ju - jin kim , phys .
lett . * 90 * , 166403 ( 2003 ) .
w. yi , l. lu , h. hu , z. w. pan , and s. s. xie , phys .
91 * , 076801 ( 2003 ) .
z. zhang , v. chandrasekhar , d. a. dikin , and r. s. ruoff , europhys . lett . * 68 * , 713 ( 2004 ) .
x. blase , l.x .
benedict , e.l .
shirley , and s.g .
louie , phys .
lett . * 72 * , 1878 ( 1994 ) .
r. landauer , philos . mag . * 21 * , 863 ( 1970 ) ; m. bttiker , phys .
lett . * 57 * , 1761 ( 1986 ) .
mintmire , d.h .
robertson , and c.t .
white , j. phys .
solids * 54 * , 1835 ( 1993 ) .
solomons , _ organic chemistry _
( wiley , new york , 1996 ) .
choi , j. ihm , s.g .
louie , and m.l .
cohen , phys .
lett . * 84 * , 2917 ( 2000 ) .
h. kim , j. lee , s .- j .
kahng , y .- w . son , s.b .
lee , c .- k .
lee , j. ihm , and y. kuk , phys .
* 90 * , 216107 ( 2003 ) .
a. silva , y. oreg , and y. gefen , phys .
b * 66 * , 195316 ( 2002 ) . t .- s .
kim and s. hershfield , phys .
b * 67 * , 235330 ( 2003 ) .
lee and c.s .
kim , phys .
b * 63 * , 75306 ( 2001 ) . c. caroli , r. combescot , p. nozieres , and d. saint - james , j. phys .
c * 4 * , 916 ( 1971 ) .
s. hershfield , j. h. davies , and j. w. wilkins , phys .
lett . * 67 * , 3720 ( 1991 ) .
y. meir and n. s. wingreen , phys .
lett . * 68 * , 2512 ( 1992 ) .
figure 4 : ( color ) .
conductance and wavefunctions for l=13 . in ( a )
, the full numerical calculation for @xmath4 ( solid red line ) is indistinguishable from the fit with the multiple fano resonance formula ( dashed red line ) . in ( b ) , the red and blue colors indicate @xmath57 signs and the size of the spheres is the amplitude of the wavefunction . | we investigate the electron transport in multiply connected metallic carbon nanotubes within the landauer - bttiker formalism .
quasibound states coupled to the incident @xmath0 states give rise to energy levels of different widths depending on the coupling strength .
in particular , donorlike states originating from heptagonal rings are found to give a very narrow level .
interference between broad and narrow levels produces fano - type resonant backscattering as well as resonant tunneling . over a significantly wide energy range ,
almost perfect suppression of the conduction of @xmath0 electrons occurs , which may be regarded as filtering of particular electrons ( @xmath1-pass filter ) . |
a new generation of radio telescopes will map the polarization of cosmic radio sources over a large range of wavelengths , from a few centimetres to several metres .
since the plane of polarization of a linearly polarized wave is rotated by an amount that depends on the magnetic field and free - electron distributions and the wavelength ( @xmath4 ) , the resulting data will probe both the synchrotron - emitting sources and any intervening magneto - ionic medium in unprecedented detail .
a useful way to characterize the intrinsic properties of magneto - ionic media is the faraday dispersion function , @xmath2 , which contains information on the transverse orientation of the magnetic field ( @xmath5 ) and on the intrinsic polarized emission as a function of faraday depth , @xmath1 .
the faraday depth is proportional to the integral along the line of sight @xmath6 of the product of the density of thermal electrons , @xmath7 , and the component of the magnetic field parallel to the line of sight : @xmath8 hence , in principle , @xmath2 can be used to obtain both the perpendicular and the parallel components of the three - dimensional magnetic field .
( our system of coordinates is such that the origin is at the far end of the source and the observer is located at @xmath9 .
a magnetic field pointing towards the observer yields a positive faraday depth . )
reconstruction of @xmath2 is usually done by taking advantage of the fourier - transform type relationship between the observed polarized emission and the faraday dispersion function .
the _ observed _ complex polarization @xmath10 can be expressed as the integral over all faraday depths of the _ intrinsic _ complex polarization @xmath2 modulated by the faraday rotation @xcite : @xmath11 so that @xmath2 can be expressed in a similar way : @xmath12 @xmath2 is a complex - valued function : @xmath13 where @xmath14 is the fraction of polarized flux that comes from regions of faraday depth between @xmath1 and @xmath15 , @xmath16 is the intrinsic polarization angle ( perpendicular to the transverse component of the magnetic field , @xmath17 ) and may itself depend on @xmath1 .
equation ( [ eqf ] ) lies at the heart of methods to recover @xmath2 from multi - frequency observations of the complex polarized intensity ( called rotation measure , rm , synthesis ; ) .
the rm synthesis has been used to recover faraday components of compact sources ( e.g. @xcite ) and diffuse structures in the milky way ( e.g. ) , in nearby galaxies ( e.g. ) and in galaxy clusters ( e.g. ) .
several techniques have been proposed to deal with the limited @xmath18 coverage provided by real telescopes ( rm - clean ; @xcite ; sparse analysis and compressive sensing ; , @xcite ; multiple signal classification ; @xcite ) and with the missing negative @xmath18 ( e.g. using wavelet transforms ; @xcite , @xcite ) .
also used wavelets to analyze the scales of structures in faraday space and emphasized the need to combine data at high and low frequencies .
because of the difficulty of the rm synthesis technique to recover multiple faraday components , it has been suggested to use direct @xmath19 and @xmath20 fitting , where @xmath21 and @xmath22 are the @xmath23 and @xmath24 stokes parameters normalised to the total intensity @xmath25 ( @xcite ; @xcite ) . in this paper we show how observations , performed in the various wavelength ranges available at existing and planned radio telescopes , can be used to constrain the variation of @xmath0 ( and therefore the orientation of the magnetic field component perpendicular to the line of sight ) with @xmath1 .
we use a fisher matrix analysis to quantify the precision that can be achieved for fitted parameters and investigate the degeneracies that exist between the different constituents of our model .
recently , @xcite performed a similar analysis to evaluate the capability of new radio telescopes to constrain the properties of intergalactic magnetic fields through observations of background polarized sources .
their work assumed two faraday components , each with a constant @xmath0 , a narrow one ( the compact radio source ) and a broad one ( possibly associated with the milky way ) .
here we consider _ a linear variation of @xmath0 with @xmath1 _ and show how the degeneracies between pairs of model parameters can be broken using complementary datasets from different instruments in order to recover @xmath3 , using two simple models of @xmath2 , a constant and a gaussian . in the simple cases we consider , the variation of @xmath26 can be produced by a helical magnetic field .
magnetic helicity is a natural consequence of dynamo action and sophisticated statistical methods have been devised to try to infer its presence , although without inclusion of faraday effects ( , ) .
anomalous depolarization ( an increase rather than the usual decrease of the degree of polarization with wavelength ) produced by an helical field was discussed by @xcite .
helical fields have been invoked to explain the anomalous depolarization properties of the nearby galaxy ngc 6946 ( ) and polarization characteristics of the central part of the starburst galaxy ngc 253 ( ) .
helical magnetic fields are also important in galactic and protostellar jets ( e.g. @xcite , @xcite ) .
bi - helical fields ( with opposite signs of helicity on small and large scales ) are produced in simulations of galactic dynamos and the signatures of such fields are discussed in a recent paper by @xcite . in this paper , we focus on the detectability of single - helical magnetic fields .
we consider observations of the stokes parameters @xmath23 and @xmath24 with the instruments listed in table [ tab1 ] .
we used a nominal integration time of 1 h for the low - frequency observations ( giant meterwave radio telescope , gmrt , westerbork synthesis radio telescope , wsrt , low frequency array , lofar ) and 10 min for observations with the more sensitive instruments ( jansky very large array , jvla , australian square kilometre array pathfinder , askap and square kilometre array 1 , ska1 ) .
this allows an easy comparison of the sensitivities and makes it possible to display the confidence intervals of the parameters of interest on a common graph ( figs [ fig2 ] and [ fig3 ] ) .
we used a channel width of 1 mhz for all instruments except the jvla for which a channel width of 2 mhz is more than sufficient in the wide s - band ( 20004000 mhz ) to resolve the main features of the spectral energy distribution of the polarization . note that all instruments allow the use of narrower channels ; however , there is an obvious trade - off between sensitivity per channel and total integration time .
we have varied the channel width over two orders of magnitude between 0.1 and 10 mhz and observed that the resulting precision on the main parameter of interest , @xmath27 ( equation ( [ eqbeta ] ) ) , changes by less than @xmath28 for a same total integration time of the ska1-survey .
the quoted sensitivities are indicative since several instruments listed in table [ tab1 ] are still in their design phase .
also , some bands , especially the low - frequency ones , will be affected by radio frequency interferences and a fraction of the channels will be missing . with real data at hand it will be straightforward to include the actual frequency coverage and sensitivities in the modeling of a particular data set .
we scaled the sensitivities @xmath29 quoted in the literature for a given effective bandwidth @xmath30 and integration time @xmath31 to new values of the channel width @xmath32 and integration time @xmath33 , as given in the table , for a given number of tunings @xmath34 to cover the whole bandwidth : @xmath35 the jvla will be used to carry out sensitive surveys of large parts of the sky .
we use figures provided by steven t. myers [ national radio astronomy observatory ( nrao ) ] in the karl jansky vla sky survey prospectus ( see table [ tab1 ] ) for the jvla in its b - configuration . in the s - band ( 24 ghz ) ,
the effective bandwidth ( free from radio frequency interferences ) is 1500 mhz , and a noise level of 0.1 mjy can be achieved in 7.7 seconds .
the size of the synthesized beam is @xmath36 at the centre of the band . in the l - band ( 12 ghz ) ,
the effective bandwidth is 600 mhz and a noise level of 0.1 mjy can be achieved in 37 seconds of integration .
the size of the synthesized beam is @xmath37 . in both cases we assumed a single tuning .
we note that @xcite recently submitted a science white paper for a jvla sky survey in the s - band ( in the c - configuration ) , in which they consider several alternatives ranging from a shallow all - sky survey to ultra - deep fields of a few tens of square degrees .
they estimate that a shallow all - sky survey of a total of about 3000 hours would lead to the detection of over @xmath38 polarized sources .
according to the askap website , a continuum sensitivity of 29 to 37 @xmath39jy beam@xmath40 for beams between 10@xmath41 and 30@xmath41 can be reached in 1 hour for a bandwidth of 300 mhz .
four tunings would be required to cover the whole frequency band from 700 to 1800 mhz , so in a total of 1 h a noise level of 1 1.3 mjy per 1 mhz channel would be reached .
a major polarization survey with askap ( possum ) is in the design study phase . in its first phase
, the ska will observe at low frequencies ( 50 350 mhz , ska1-low ) , mid - frequencies ( 0.35 3.05 ghz , ska1-mid ) and in a survey mode in the 0.65 1.67 ghz range ( ska1-survey ) . in 1 hour of observation and per 0.1 mhz channel , the sensitivity is expected to be 63 @xmath39jy for ska1-mid , 103 @xmath39jy for ska1-low , and 263 @xmath39jy for ska1-survey . we have assumed a single tuning for ska1-low and that for ska1-mid four tunings ( in a 770 mhz bandwidth each ) will be needed to cover the whole band ; for ska1-survey , the maximum bandwidth will be 500 mhz , so two tunings will be needed .
the corresponding noise levels per 1 mhz bandwidth and after 10 minutes of observations are given in table [ tab1 ] .
for the gmrt 610 mhz band , our sensitivity estimate is based on the figures quoted by @xcite who reached a noise level in @xmath23 and @xmath24 of 36 @xmath39jy per beam of 24@xmath41 in 180 minutes in a 16 mhz band centered at 610 mhz .
four tunings would be required to cover the whole band .
our estimate of the sensitivity of the gmrt in the 325 mhz band relies on a noise level of 2.7 mjy per beam per 1 mhz channel in 1 hour , based on polarization observations of a pulsar done in 2011 ( farnes , private communication ) and assuming that all 30 antennas would be available . assuming that 2 tunings would be necessary to cover the whole band , this gives a noise level of 3.8 mjy in a total of 1 h. the wsrt also operates in the 320 mhz band ( called the 92 cm band ) .
after 1 hour of observation , the theoretical noise level in stokes @xmath25 is 1.2254 mjy beam@xmath40 in a 10 mhz band .
this corresponds to about 3.9 mjy beam@xmath40 in a 1 mhz channel .
note that confusion noise is expected to be significant in observations of the stokes parameter @xmath25 but it can be neglected @xmath23 and @xmath24 . recently detected polarization with the wsrt towards the andromeda galaxy at 350 mhz .
the high - band array ( hba ) of the lofar operates at frequencies between 110 mhz and 250 mhz with a filter between 190 and 210 mhz .
lofar has detected polarization in the hba and rotation measures could be inferred ( e.g. in pulsars , , and in polarized sources in the field of m 51 , mulcahy et al . , in prep . ) . however , at the lofar frequencies depolarization is extremely strong and for the fiducial models presented in this paper the measurements at the quoted sensitivities do not provide improved constraints on the parameters related to the magnetic field . the lofar frequency coverage is displayed in fig . [ fig1 ] but
the lofar confidence intervals are therefore not shown in the other figures .
the fisher analysis is often used in cosmology ( e.g. @xcite , page 94 ) . consider a set of @xmath42 data points and a model with @xmath43 parameters , @xmath44 .
the fisher matrix elements @xmath45 are proportional to the second partial derivatives with respect to two given parameters of the likelihood function @xmath46 that the data set derives from the given model . if the measurement errors follow a gaussian probability distribution , then @xmath47 where @xmath48 is defined in eq .
( [ eqchi2 ] ) . denoting @xmath49 and @xmath50 the values of the stokes parameters @xmath23 and @xmath24 for the assumed model , estimated at wavelengths @xmath51 and with noise levels @xmath52
, we have @xmath53 the fisher matrix elements can be written as @xmath54 the covariance matrix is the inverse of the fisher matrix : @xmath55 we consider two simple models for @xmath2 , each with a linearly varying @xmath0 as a function of faraday depth : @xmath56 with some constants @xmath57 and @xmath27 .
this is a parametrisation of @xmath16 as a first - order polynomial and , as discussed below , it can also be interpreted as a helical magnetic field .
one of the simplest possible models for @xmath2 is the top - hat , @xmath58 where the set of parameters is @xmath59 , @xmath60 is given by eq .
( [ eqpsi0 ] ) and @xmath61 is the top - hat function , with @xmath62 in the range @xmath63 and @xmath64 elsewhere .
the complex polarization is @xmath65 e^{2{\rm i } \psi(\lambda^2 ; { \bf{p } } ) } \ , , \label{eqptophat}\ ] ] where @xmath66 and @xmath67 . in a uniform slab ,
the faraday depth varies linearly with the @xmath6-coordinate for @xmath1 between @xmath68 : @xmath69 where @xmath7 is in @xmath70 , @xmath71 in @xmath39 g , @xmath6 in pc and @xmath1 in rad m@xmath72 .
consider a magnetic field with a constant line - of - sight component , but with a rotating component in the plane of the sky : @xmath73 the intrinsic polarization angle clearly varies with the faraday depth : @xmath74 where @xmath75 for an helical field with @xmath76 rad kpc@xmath40 , we have @xmath77 m@xmath78 .
note that the sign of @xmath27 depends on the relative orientation of the magnetic field component along the line of sight and the handedness of the helix .
a positive @xmath27 means that @xmath71 , which produces the faraday rotation , and the intrinsic rotation of @xmath79 have the same direction .
a negative @xmath27 means that the faraday rotation effectively counteracts the intrinsic rotation of the plane - of - sky magnetic field .
this effect will be discussed further in sect .
3 . as a simple alternative to the top - hat parametrisation of @xmath2
we also consider a gaussian form , @xmath80 e^{2{\rm i } \psi_0(\phi ; { \bf p})},\ ] ] where @xmath81 and @xmath82 are defined as before .
this gives a complex polarization of @xmath83 e^{2{\rm i } \psi(\lambda^2 ; { \bf p } ) } , \label{eqpgaussian}\ ] ] where @xmath84 is given by eq .
( [ psiobs ] ) .
the modulus of @xmath10 is a gaussian centered at @xmath85 with variance @xmath86 . in the two previous sections we calculated @xmath10 by integrating eq .
( [ eqpol ] ) analytically .
using the properties of the fourier transforms , we now show why any linear variation of @xmath0 with @xmath1 produces a translation of the observed polarized intensity in the @xmath18 space . using the standard expression for fourier transform ( integral over @xmath87 from @xmath88 to @xmath9 of a function @xmath89 times @xmath90 for the direct transform , and times @xmath91 for its inverse ) , eq .
( [ eqpol ] ) can be written as @xmath92 where @xmath93 is the inverse fourier transform . using @xmath94 where @xmath95 is a real - valued function centered at @xmath96 , @xmath97 the factor @xmath98 is independent of @xmath1 and
can be taken out of the integral .
multiplication becomes a convolution in the fourier ( @xmath18 ) space and convolution becomes a multiplication , so @xmath99 translation in the @xmath1-space gives a rotation in @xmath18-space , and the inverse transform of the term involving @xmath100 becomes a delta - function , giving @xmath101 we then obtain @xmath102 where @xmath103 is the complex polarization corresponding to @xmath95 when @xmath104 . changing the variable from @xmath105 to @xmath18 , we obtain @xmath106 so that @xmath107 this shows that the observed modulus of the polarized intensity of a medium with a given @xmath27 is simply a translation by @xmath27 in @xmath18-space of what would be observed if @xmath27 were equal to zero .
@xmath108 is real - valued , by definition ; if it is even in @xmath1 , its inverse fourier transform is also real - valued , and the observed polarization angle will be @xmath109 as found in sect .
[ sect231 ] and [ sect232 ] for the top - hat and the gaussian cases . .
the askap regions are in blue and the gmrt ones in green .
, scaledwidth=48.0% ]
we use a fiducial top - hat model of the faraday dispersion function with the following parameters : @xmath110 = 15 rad m@xmath72 , @xmath111 rad m@xmath72 , @xmath112 = 0.1 mjy/(rad m@xmath72 ) , @xmath113 rad , and three values of @xmath27 , 0 and @xmath114 m@xmath78 . the total intrinsic polarized flux density ( integrated over all faraday depths )
is thus @xmath115 mjy .
because of depolarization , the signal expected at a given frequency will be weaker ( see figure [ fig1 ] ) but should be detectable within a reasonable amount of observing time by current and future instruments ( see table [ tab1 ] ) . for comparison , we also use a gaussian model of @xmath116 with the same total flux .
the dispersion of the gaussian is @xmath117 rad m@xmath72 and the peak flux density per unit of faraday depth @xmath118 mjy/(rad m@xmath72 ) .
note that @xmath119 characterises the gaussian profile of the faraday dispersion function for a model with a regular field , not to be confused with the dispersion in rotation measure ( rm ) caused by possible rm fluctuations across an observing beam ( usually denoted @xmath120 ) . in sect .
[ sectfaradaydispersion ] we discuss the additional depolarization effect by faraday dispersion produced by a random field .
the width of the faraday structure ( that is , the total faraday depth , sometimes denoted @xmath121 , see sect [ sectfaradaydispersion ] ) in our fiducial top - hat model is @xmath122 rad m@xmath72 . in many astrophysical cases
this quantity can be larger ( up to @xmath123 rad m@xmath72 in spiral arms , e.g. @xcite ) .
a larger total faraday depth translates into a narrower main peak " of the @xmath10 distribution and weaker emission at long wavelengths .
[ fig1 ] shows the variation of the polarized intensity with @xmath18 for a top - hat ( left column ) and a gaussian ( right column ) faraday dispersion function . in the former case , the solid line ( @xmath124 ) is the well - known sinc function produced by a uniform slab , which is more usually shown using the linear horizontal axis used in the bottom left panel .
note that , for clarity , the graphs in fig .
[ fig1 ] do not include any spectral dependence of the intrinsic polarization .
figures [ fig2 ] and [ fig3 ] ( the confidence regions of the parameters ) do , on the other hand , include a spectral dependence of the form given by eq .
( [ eqspecdep ] ) .
in the gaussian case , the polarized intensity decreases monotonically and no emission is produced in the longer wavebands at which the gmrt and lofar operate ( fig .
[ fig1 ] ) . in the bottom right panel
we show the variation of the @xmath23 and @xmath24 stokes parameters ( in brown and blue ) for the gaussian @xmath2 , which are the direct observables .
they oscillate with a 90@xmath125 phase shift with respect to each other . in the rest of the paper we focus on the top - hat model because it gives stronger emission in longer wavebands for the parameters selected here and it includes the standard case for faraday depolarization calculations of a uniform slab as a special case .
it is most interesting to compare the variations of the polarized intensity for a positive and a negative @xmath27 . when @xmath126 , the intrinsic helicity of the magnetic field and the faraday rotation act in the same direction .
this results in an increased depolarization at short wavelengths . even for @xmath127 , where faraday depolarization is absent ,
the emission is significantly depolarized compared to the case of a constant magnetic field orientation ( @xmath124 ) . when @xmath128 , faraday rotation counteracts the intrinsic rotation of the field , which means that the polarized emission peaks at a wavelength different from zero ( dashed lines ) , where @xmath85 ( eqs .
( [ eqptophat ] ) and ( [ eqpgaussian ] ) ) .
this effect is similar to the `` anomalous depolarization '' discussed by ( * ? ? ?
* section 9 ) .
[ fig2 ] shows the 68.3% confidence regions of the five different parameters obtained from the fisher analysis for @xmath124 .
the colour code is the same as in fig .
[ fig1 ] , with the askap confidence regions in blue and the gmrt ones in green . the plots in the first row show that the central faraday depth @xmath110 is mostly uncorrelated with the other parameters .
on the other hand , @xmath129 , which describes the extent of the faraday component in @xmath1-space , is strongly correlated with the parameters related to the intrinsic polarization angle ( @xmath57 and @xmath27 ) and at short wavelengths ( in the askap band ) with the normalization of @xmath2 ( @xmath112 ) .
this is because the crucial effect is faraday _ differential _ rotation across the faraday component and not the magnitude of the central faraday depth .
an increase in @xmath129 means stronger depolarization due to differential rotation which must be counteracted by a more negative @xmath27 in order to produce a similar fit to the data .
vice versa , a lower @xmath129 means weaker depolarization and @xmath27 needs to become positive to increase the depolarization . in other words , @xmath129 and @xmath27 are anti - correlated around @xmath124 .
most importantly , the strength of the correlation between pairs of parameters including @xmath57 or @xmath27 varies with the selected waveband .
this is what makes it possible to break parameter degeneracies by combining short - wavelength ( like askap ) and long wavelength ( like gmrt ) data sets in order to better constrain the derived parameter values , as we discuss next .
[ fig3 ] shows the one - sigma confidence ellipses for @xmath57 , @xmath27 , @xmath112 and @xmath129 for @xmath130 m@xmath78 ( four left panels ) and @xmath131 m@xmath78 ( four right panels ) .
as expected , the orientation of the ellipses is similar for the short - wavelength instruments jvla ( s - band in purple and l - band in pink ) , askap ( blue ) and ska1-survey ( black ) .
on the other hand , the long - wavelength gmrt data set ( green ) produces a different correlation between parameters ; in some cases the confidence ellipses at short and long wavelengths are almost orthogonal , making it possible to reduce the confidence intervals on the derived parameter values considerably by using both wavelength ranges together ( such as askap and gmrt , shown in yellow ) .
this is further illustrated in fig .
[ fig4 ] where the precision that can be achieved in @xmath27 for different instruments is shown .
a combination of askap and gmrt observations makes it possible to reach an uncertainty @xmath132 m@xmath78 if @xmath128 , for the integration times shown in table [ tab1 ] and the set of model parameters used .
if @xmath126 , all signals are weaker because of increased depolarization and the precision on @xmath27 ( and all other model parameters ) is lower .
we examined how variations of the intrinsic polarization angle @xmath0 with the faraday depth @xmath1 within a source affect the observable quantities . using simple models for the faraday dispersion @xmath2 and @xmath3 , along with the current and planned properties of the main radio interferometers ,
we show how degeneracies among the parameters describing the magneto - ionic medium can be minimised by combining observations in different wavebands .
in particular we have shown that it may be possible to recover the sign and the magnitude of @xmath27 , a parameter that we have defined and that is related to the relative effect of the helicity of the transverse magnetic field and the faraday rotation due to the parallel component of the magnetic field .
since the direction of @xmath133 can be easily inferred from rm measurements , it should be possible to recover the sign ( and , under some assumptions the magnitude ) of the helicity of the magnetic field .
however , the additional effect of faraday dispersion by a random component of the magnetic field attenuates the variations of the polarized emission as a function of wavelength and may shift the peak of polarized emission towards shorter wavelengths if @xmath27 is negative .
faraday depolarization effects ( both by differential rotation and by dispersion ) will have to be included in the modelling of real data in order to recover information on the helicity of the magnetic field .
this approach is complementary to statistical studies of the correlation between the degree of polarization and the rotation measures of cosmic sources which may also provide information on magnetic helicity ( @xcite , @xcite ) .
planned surveys of fixed sensitivity will be biased towards radio sources with negative @xmath27 , because of the depolarization produced when @xmath134 .
detection of @xmath134 will be more difficult both through @xmath135 and @xmath136 model - fitting and rm synthesis because most of the signal is shifted toward negative @xmath18 .
@xcite show that restricting the integral in eq .
( [ eqf ] ) to the positive @xmath18 yields an erroneous reconstruction of @xmath2 when @xmath126 . in this work
we used a first - order parametrisation of the variation of intrinsic polarization angle with faraday depth .
higher - order representations could be used ( or e.g. chebyshev polynomials ) if the data are of sufficient quality . including a second - order term implies a convolution with an imaginary gaussian and a significantly more complicated expression for @xmath10 .
we thank oliver gressel for organizing the stimulating nordita workshop ` galactic magnetism in the era of lofar and ska ' held in stockholm in 2013 september .
we also thank axel brandenburg and rodion stepanov for interesting discussions on the topics discussed in this paper and for sharing their related results with us .
we are grateful to rainer beck , john h. black , jamie farnes , george heald , anvar shukurov and to the referee for constructive comments .
af is grateful to the leverhulme trust for financial support under grant rpg-097 . | wide - band radio polarization observations offer the possibility to recover information about the magnetic fields in synchrotron sources , such as details of their three - dimensional configuration , that has previously been inaccessible .
the key physical process involved is the faraday rotation of the polarized emission in the source ( and elsewhere along the wave s propagation path to the observer ) . in order to proceed ,
reliable methods are required for inverting the signals observed in wavelength space into useful data in faraday space , with robust estimates of their uncertainty . in this paper
, we examine how variations of the intrinsic angle of polarized emission @xmath0 with the faraday depth @xmath1 within a source affect the observable quantities . using simple models for the faraday dispersion @xmath2 and @xmath3 , along with the current and planned properties of the main radio interferometers ,
we demonstrate how degeneracies among the parameters describing the magneto - ionic medium can be minimised by combining observations in different wavebands .
we also discuss how depolarization by faraday dispersion due to a random component of the magnetic field attenuates the variations in the spectral energy distribution of the polarization and shifts its peak towards shorter wavelengths .
this additional effect reduces the prospect of recovering the characteristics of the magnetic field helicity in magneto - ionic media dominated by the turbulent component of the magnetic field .
[ firstpage ] polarization methods : data analysis techniques : polarimetric ism : magnetic fields galaxies : magnetic fields radio continuum : galaxies |
nonequilibrium systems with similar symmetries often form patterns which appear to be universal in spite of having been formed by different physical mechanisms @xcite . in particular , reduced descriptions of the patterns often quantify the similarities in behavior so that understanding of one system can lead to insights in multiple systems . a class of spatiotemporally chaotic states exhibiting defect - mediated turbulence @xcite has been found in such diverse systems as wind - driven sand , electroconvection in liquid crystals @xcite , nonlinear optics @xcite , fluid convection @xcite , and autocatalytic chemical reactions @xcite . in many cases , such systems have been modeled via the complex ginzburg - landau equation @xcite .
these various defect turbulent patterns are characterized by an underlying striped state which contains dislocations ( point defects ) where the stripes dead - end within the pattern .
locally , the defects distort the orientation and wavenumber of the stripes and the nucleation , motion , and annihilation of the defects constitute a spatiotemporally chaotic system .
an example from inclined layer convection is shown in fig .
[ d_f_pic ] .
previous work on defect turbulence has focused both on snapshots of such patterns @xcite and the dynamics and interaction @xcite .
however , there are numerous open questions about defect turbulence : characterization of the defect motions , interactions between the defects , and the extent to which the analogy to turbulence is appropriate .
the degree to which such characteristics are similar in different defect - turbulent systems remains to be explored .
investigations of pattern formation in variants of rayleigh - bnard convection ( rbc ) have been particularly fruitful @xcite .
the state of undulation chaos ( shown in fig .
[ d_f_pic ] and described in @xcite ) observed in inclined layer convection ( fig .
[ d_f_schematic ] ) exhibits defect turbulence and is well suited to investigations on the dynamics of defects since spatially extended systems and fast time scales are experimentally accessible .
this allows for tracking of point defects through their creation , motion , and annihilation . in the observed pattern ,
the stripes contain undulations as well as defects ; both are spatiotemporally chaotic ( further characterization to be published in @xcite ) .
a number of features stand out in the defect trajectories we observe in undulation chaos . as exemplified in fig .
[ d_f_defecttrack ] , the inherent anisotropy ( due to the inclination ) is apparent , with the trajectory meandering preferentially in the transverse direction . occasionally , there is a rapid burst of transverse motion as marked by the black diamonds , corresponding to a tearing of the pattern across the rolls .
such behavior appears be related to lvy flights @xcite for which distributions of step sizes display power laws .
furthermore , we are able to study the ensemble of trajectories to gain insight into defect pair creation , interaction , and annihilation .
in inclined layer convection ( ilc ) , a thin fluid layer heated from one side and cooled from the other is tilted by an angle @xmath0 ; the system is anisotropic due to the resulting shear flow ( see fig .
[ d_f_schematic ] ) .
the fluid becomes unstable above a critical temperature difference @xmath1 . at fixed @xmath0 ,
we describe how far the system is above the onset of convection via the non - dimensional driving parameter @xmath2 . at low angles of inclination
, buoyancy provides the primary instability ( analogous to rbc ) and the convection rolls are aligned with the shear flow ( longitudinal ) both at onset and for the secondary transition to undulation chaos @xcite . over a range of intermediate angles ( @xmath3 for prandtl number @xmath4 ) ,
the longitudinal rolls become unstable to undulation chaos above @xmath5 .
it is this defect turbulent state which we investigate ; some of its properties have been previously described in other work @xcite . above @xmath6 , there is a further transition to a state of competing ordered undulations and undulation chaos .
we examine trajectories from both sides of this transition .
the apparatus used in this experiment is of the type described in de bruyn et al .
@xcite , modified to allow for inclination .
the fluid used was co@xmath7 at a pressure of @xmath8 bar regulated to @xmath9 bar with a mean temperature of @xmath10c regulated to @xmath11c . as determined from a materials properties program @xcite , the prandtl number was @xmath12 .
a cell of height @xmath13 m and dimensions @xmath14 was used , for which the vertical diffusion time was @xmath15 sec .
the fluid was weakly non - boussinesq conditions : @xmath16 , as described in @xcite for horizontal fluid layers .
all experiments were performed at a fixed inclination of @xmath17 , within the regime of buoyancy - instability .
images of the convection pattern were collected using a digital ccd camera , via the usual shadowgraph technique @xcite .
images were collected at 3 frames per second in one of two formats .
six - hour ( @xmath18 , 80000 frames ) continuous runs of data were obtained at two values of @xmath19 : 0.08 ( four runs ) and @xmath20 ( two runs ) . for 17 values of @xmath19 between 0.04 and 0.22 , short runs with 100 images were collected , separated by at least @xmath21 for statistical independence . at each of these @xmath19 ,
at least @xmath22 repeats ( up to 600 for the lowest values of @xmath19 ) were performed .
each value of @xmath19 was reached by a quasistatic temperature increase from below .
in addition , a run with quasistatic temperature decreases was performed between @xmath23 and @xmath24 to check for hysteresis , which was not observed .
only data from the central homogeneous region of dimension @xmath25 was utilized during the analysis unless noted otherwise ; see @xcite for details on the choice of this region .
size - dependent effects of this region are discussed here .
here we consider only the defects themselves an ensemble of moving , interacting , charged `` particles '' and not the underlying pattern . by reducing the data to a set of defect trajectories of known charge , birth , and death
we can observe how these particles move individually and as an ensemble .
single defects may enter or leave through the edges of the system , or be created / annihilated in pairwise events between defects of opposite topological charge @xcite . during their lifetimes , the defects have positions and velocities which may be related both to the underlying pattern and to the presence of other defects .
we examine properties averaged over these effects , and also isolate some effects due to the latter .
the topological defects are located at points where a roll pair ends within the pattern ( see fig .
[ d_f_pic ] ) . at these points
, there is a phase discontinuity and the topological charge is determined from the phase jump along a contour around the defect : @xmath26 the phase field @xmath27 is determined from fourier - demodulated images for which the complex field @xmath28 has been reconstructed .
we detect the locations of defects by finding points where @xmath29 and @xmath30 and a phase discontinuity occurs .
( details of this technique are described in @xcite . ) in undulation chaos , we observe only defects with @xmath31 ; examples of each topological charge are marked by circles in fig .
[ d_f_pic ] .
after the defects were detected in each image , the defect locations were connected to form trajectories .
this was done by matching each defect already on a trajectory to its closest same - signed neighbor in the subsequent frame .
if the reverse process also agreed on the match , then the subsequent defect was added to the list of defects located on that trajectory .
similarly , the ends of positive and negative defect trajectories were matched up to obtain the locations of all pair creations and annihilations .
when the analysis was completed , broken trajectories ( those missing a creation or annihilation ) were rejoined .
since the defect positions were determined only to the nearest pixel , the data was later smoothed to interpolate the coordinates along the trajectory .
each @xmath32 and @xmath33 along a trajectory was replaced with the weighted average of its neighbors , using the gaussian weighting function @xmath34 \\ & \cdot \exp [ { - \frac{(y_i - y_j)^2}{2 \sigma^2_y } } ] \ ,
\cdot \ , \exp [ { - \frac{(t_i - t_j)^2}{2 \sigma^2_t } } ] \label{d_e_weight}\end{aligned}\ ] ] this resulted in a smoothed trajectory with coordinates @xmath35 and @xmath36 ( tildes will be dropped hereafter for convenience ) .
the appropriate widths @xmath37 in @xmath38 were determined by examining a range of parameters and finding convergence for @xmath39 and @xmath40 .
the advantage of this weighting method is that it automatically adjusts the fit length so that along faster - moving segments of the trajectory we average fewer data points than along slower - moving segments . from the trajectories
@xmath41 and @xmath42 , we applied a local weighted linear fit ( again using @xmath38 ) at each point along the trajectory to obtain the corresponding velocities @xmath43 and @xmath44 .
we also determined @xmath45 and @xmath46 , corresponding to a downslope ( @xmath47 ) direction at @xmath48 . a raw trajectory
is shown in fig .
[ d_f_defecttrack ] , along with the smoothed trajectory and the transverse velocity component @xmath43 . defect turbulence in inclined layer convection is anisotropic , with @xmath49 corresponding to glide motion across the rolls and @xmath50 to climb motion along the rolls .
these are fundamentally different behaviors : the former adjusts the orientation of the rolls , and the later the wavenumber . as expected , the decomposed motions are found to be poorly correlated with each other , with a linear correlation coefficient @xmath51 for positive defect trajectories and @xmath52 for negative . therefore , we separate all defect positions and motions into their two components and examine each direction independently .
defect velocities are observed to be correlated over time scales of less than @xmath53 , as determined from temporal autocorrelation functions of @xmath43 and @xmath44 .
the autocorrelations are plotted in fig .
[ d_f_velautocorr ] , averaged over the ensemble of trajectories .
the defect motions exhibit short - range order , with exponential decay during the time before the zero - crossing .
we characterize the meanderings of the defect via its central moments , represented by integer powers of the deviations from its mean value . for trajectories
@xmath54 with mean @xmath55 , the @xmath56 moment is given by the quantity @xmath57 where @xmath58 represents an ensemble average , and a similar calculation can be made for the @xmath42 . for normal diffusion or random
walks , @xmath59 , where @xmath60 provides a value @xmath61 for the diffusion constant .
investigations in a broad variety of situations ( see , for example , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) have found anomalous diffusion instead , where @xmath62 . for @xmath63
a system is said to be subdiffusive , @xmath64 is superdiffusive , and @xmath65 is ballistic transport .
anomalous diffusion is frequently associated with lvy walks / flights , in which the pdfs of the lengths and durations of the flights are power laws and have infinite moments .
since the central limit theorem no longer holds in such cases , the probability of long jumps ( waits ) will enhance ( retard ) the diffusive behavior .
similarly , the power spectrum @xmath66 of the associated velocities will exhibit power law behavior as well .
some recent examples of such behavior in fluid dynamics include tokamak density fluctuations @xcite and particle motion in two - dimensional rotating fluid flows @xcite .
defect random walks have also been considered for models of intracellular ca@xmath67 dynamics @xcite , in which subdiffusion was observed . in fig .
[ d_f_diff8 ] and [ d_f_diff17 ] we calculate moments for those values of @xmath68 at which 100 or more trajectories have at least that duration .
we look for behavior of the type @xmath69 , @xmath70 , and @xmath71 in both the @xmath72 and @xmath73 directions . in the variance ( @xmath74 ) ,
superdiffusive behavior is evident at both @xmath19 , particularly in @xmath75 where @xmath76 .
the exponents for @xmath47 are lower but also indicate superdiffusion .
the exponents of @xmath77 and @xmath78 do not suggest superdiffusion in @xmath72 , and the @xmath47 behavior does not follow a power law .
we locate defects flights as the portions of the trajectories which occur while @xmath79 . an example flight is shown in fig . [ d_f_defecttrack ] . for each of these flights
, we determine the duration @xmath80 and displacement @xmath81 .
distributions of these values are shown in fig .
[ d_f_flights ] . due to the constraints of a finite system size
, the maximum observable @xmath81 is limited .
nonetheless , the distribution of flight times @xmath80 is observed to have behavior consistent with a power law tail of exponent approximately @xmath82 . for @xmath81 , behavior suggestive of a power law with exponent approximately @xmath83
is observed , indicating that there is not a characteristic velocity associated with the flights .
physically , these flight events correspond to a tearing of the rolls when the pattern shifts along a line in the @xmath72 direction , so that each roll broken in the process rejoins with the one next to it .
such a tearing line is visible next to the white arrow in fig .
[ d_f_pic ] , where there is weaker convection along the slopes of the undulations .
since each roll must move by only @xmath84 to reconnect with the next roll , defects can be transported rapidly from one end of the tear line to the other . by analogy to fluid turbulence ,
we plot probability distribution functions ( pdfs ) and power spectra of the defect velocities . figs .
[ d_f_veldist8 ] and [ d_f_veldist17 ] show the results for distributions in @xmath72 and @xmath47 .
the pdfs of @xmath49 are independent of topological charge and show steep power law tails with an exponent of approximately @xmath85 .
this is consistent with the results obtained for defect flights displacements and durations . as can be seen in figs .
[ d_f_veldist8 ] and [ d_f_veldist17 ] , the peak of the distributions have a gaussian shape , indicative of smoothing due to noise .
each distribution is also of a shape described by tsallis statistics of the type described in @xcite .
the pdfs for @xmath50 show lower characteristic velocities and ( particularly at @xmath86 ) dependence on the topological charge .
[ d_f_velanglecorr ] shows a complementary picture in @xmath87 space via a two dimensional histogram .
the @xmath72 flights are now visible as the ridges at approximately @xmath88 .
this transverse direction not only comprises most of the defect motion , but is also the direction along which motion is the fastest . @xmath89 @xmath90 a second effect visible in figs .
[ d_f_veldist8 ] , [ d_f_veldist17 ] , and [ d_f_velanglecorr ] is an asymmetry in the @xmath91 behavior which shifts the @xmath92 peaks in the upslope or downslope direction , away from @xmath93 . although the @xmath94 defects are more isotropic in their motion than the @xmath86 defects , in both cases the negative defects prefer the downslope direction over the upslope , and the positive defects prefer the upslope .
in fact , figs . [ d_f_veldist8 ] and [ d_f_veldist17 ] show that this asymmetry is the dominant behavior at lower @xmath19 : it is rare for defects to travel other than in the preferred direction .
asymmetries in defect behavior with respect to topological charge have been observed in other anisotropic systems such as langmuir circulations @xcite , although for quantities such as the surface convergence rather than the defect velocity . in this case , the asymmetry may be related to the breaking of the @xmath95 symmetry by non - boussinesq effects . in a system consisting of lvy flight behavior
, the power spectra @xmath66 of @xmath96 is expected to show power law behavior for @xmath97 , where @xmath98 is a short time associated with the low-@xmath68 limit of power law behavior of the pdf of @xmath80 for flights @xcite .
such a cutoff is relatively large ( @xmath99 ) in the defect flight durations shown in fig .
[ d_f_flights ] , limiting the range over which we would expect to see power laws in @xmath66 for defect trajectories .
[ d_f_spectrum ] suggests random - walk behavior ( constant , uniform spectrum ) for low @xmath100 , crossing over to a flight - related power law at high @xmath100 .
the exponent in all cases ( @xmath19 , @xmath72 , @xmath47 , and topological charges ) is approximately @xmath101 .
the defects carry topological charge , and the resulting phase fields allow them to interact with each other . by examining the relative position of nearby defects , we can gain insight into their interactions .
we examine strips of width @xmath102 ( or @xmath103 ) and calculate the distances @xmath104 ( or @xmath105 ) separating all pairs of defects within the strip ( see fig .
[ d_f_strippic ] for a schematic ) .
[ d_f_stripsep ] plots pdfs of this data for all @xmath106 , @xmath107 , and @xmath108 pairs of defects : @xmath109 .
the data is limited by the finite size of the region under consideration , and is only plotted over half the dimensions of the region . for both @xmath19 ,
it is rare for oppositely - charged defects to be located within @xmath110 of each other .
this is associated with the annihilation of such pairs which pass too close in either direction .
defect pairs of @xmath107 or @xmath108 charge behave in similar fashions to each other , demonstrated by the closely overlapping dashed and dotted lines of fig .
[ d_f_stripsep ] .
there is a repulsive effect in the @xmath72 direction , with @xmath104 increasing away from 0 .
the turnover in the @xmath94 pdf may be due to finite size considerations . in the @xmath47 direction
, there is again an excluded region of around @xmath111 , an effect which is strong enough that secondary and tertiary peaks at @xmath112 and @xmath113 are visible as well for @xmath94 .
a general feature of defect turbulent systems is the production and loss of defects through two mechanisms : pair creations / annihilations and flux through the boundaries .
a mean - field approximation has been postulated by gil et al . @xcite and
experimentally examined by falcke et al .
@xcite and daniels and bodenschatz @xcite in which these rates depend on the number of defects in the system .
@xmath114 where @xmath115 is the probability of a pair creation event happening per unit time , @xmath116 is pair annihilation , @xmath117 is a single defect leaving , and @xmath118 is a single defect entering .
@xmath119 is taken to be either the number of positive defects @xmath120 or the number of negative defects @xmath121 , quantities which are approximately equal on average .
[ d_f_netcharge ] shows pdfs of the net charge @xmath122 for the observed images , with a mean of zero ( topologically neutral ) , independent of @xmath19 . the constant creation and entering rates
can be understood as being generated by random events , independent of the number of defects already in the system .
the annihilation rate scales as the number of positive defects times the number of negative defects .
the leaving rate is proportional to the number of defects present in the system .
based on this approach , we earlier derived and tested a universal distribution for @xmath119 which agrees with the experimental findings @xcite .
we find , as well , that the subscripted coefficients depend on @xmath19 , system size , and other physical parameters in systematic ways .
the stationary distribution for @xmath119 was found using a recursive relation @xmath123 to describe the probabilities at adjacent @xmath119 @xcite .
in fact , the more stringent condition of detailed balance holds as well . by
time - reversing all trajectories , creations become annihilations and entering defects leave the subregion .
analysis of the reversed trajectories for @xmath124 is shown in fig .
[ d_f_reverse ] in comparison with the original data and found to exhibit the same behavior . such defect gain and
loss rates properly scale with the size of the region under consideration .
we define a set of size - independent coefficients @xmath125 , @xmath126 , @xmath127 , and @xmath128 to quantify this behavior .
both creation and entering are random events , which have some rate per unit area and length , respectively .
thus , @xmath129 and @xmath130 where @xmath131 and @xmath132 are the surface area and perimeter of the subregion , respectively . with primed and unprimed variables representing different subregion sizes , @xmath133 and @xmath134 . for two subregions with the same density of defects @xmath135 , the annihilations per unit area ( at constant @xmath136 ) is also constant . for @xmath137 and constant density of defects
@xmath136 , @xmath138 .
analogously , @xmath139 . fig .
[ d_f_coeff_boxsize ] compares these rescaled coefficients for various sizes of test regions .
because of the @xmath72-@xmath47 anisotropy , regions with the same aspect ratio as the homogeneous subregion were used .
all four coefficients show constant rescaled rates for a given @xmath19 .
there is , however , a slight trend in the annihilation rates : larger boxes allow more defects to find each other and annihilate , particularly at higher @xmath19 .
this may also relate to the larger test regions approaching the less homogeneous sidewall regions ( see @xcite ) .
[ d_f_coeffs ] shows the coefficients @xmath140 , @xmath141 , @xmath142 , and @xmath143 as a function of @xmath19 .
the error bars were determined via a bootstrap method : by re - sampling the data with replacement we obtained a distribution of values for each coefficient from which to estimate the error . at present
we have no explanation for the particular shapes of these graphs , which appear to be linear to first order .
both @xmath144 and @xmath145 appear to intercept the @xmath19-axis close to the @xmath146 onset of undulation chaos @xcite . for @xmath147 and @xmath148
, both lines extrapolate to the origin , the expected behavior for @xmath149 .
the data at @xmath150 and 0.05 have been disregarded due to finite - size effects which are significant at low values of @xmath19 since the undulations are a long - wavelength ( @xmath151 ) instability .
defects are increasingly rare for @xmath152 as the wavelength of the undulations becomes on the order of the size of the convection cell .
the defect trajectories observed in inclined layer convection display many intriguing behaviors . the occasional rapid motion of defects across many convection rolls can be associated with power laws in various quantitative measures of their motility : diffusion , velocity pdfs , flight size pdfs , and velocity power spectra . while the system is strongly anisotropic and its defect motions are dominated by transverse motion , pdfs of the defect separations reveal significant correlations in the longitudinal direction as well .
finally , the trajectories allow us to probe the gain and loss of defects in which much of this interesting behavior can be averaged over to reveal general results .
we have focused primarily on characterizing the observed flight behavior .
if these rapid motions are , in fact , related to lvy flights then the exponents found in each of the power laws should be related to each other as has been presented in the literature for various random walk formalisms @xcite .
however , comparisons to existing theory are difficult due to both the strongly anisotropic nature and the finite size of this system
. further investigation into this phenomenon , possibly in other systems , will undoubtedly be fruitful .
the trajectories analyzed here were obtained for @xmath89 and 0.17 , of which the former is in a state of undulation chaos and the later intermittently exhibit ordered undulations .
the dynamics of this transition is discussed in @xcite , and the defect trajectories should also be examined in light of the existence of the two states .
the statistics presented here will presumably vary according to whether the system is in an ordered or disordered state . furthermore , the motion of the defects during the transition may shed light on the nature of the transition .
other questions remain regarding the relation between the defect motion and the underlying undulation pattern .
for instance , the relationship between the motion of the defects and the stability of the local wavenumber .
a related issue is that defects have been observed to `` bounce '' off regions of ordered undulations , sharply reversing direction .
finally , the relationship between the relative position and relative velocity of defects has not yet been investigated .
further work in this area would provide information about the attraction and repulsion of defects , particularly near creation and annihilation events . | we report experiments on defect - tracking in the state of undulation chaos observed in thermal convection of an inclined fluid layer .
we characterize the ensemble of defect trajectories according to their velocities , relative positions , diffusion , and gain and loss rates .
in particular , the defects exhibit incidents of rapid transverse motion which result in power law distributions for a number of quantitative measures .
we examine connections between this behavior and lvy flights and anomalous diffusion .
in addition , we describe time - reversal and system size invariance for defect creation and annihilation rates . *
pacs : * 47.54.+r , 47.20.bp * keywords : * pattern formation , convection , defect turbulence , anomalous diffusion , lvy flights * topological defects within patterns are observed in many systems to move in a spatiotemporally chaotic fashion .
we examine the motion of such defects within a defect - turbulent state observed in thermal convection of an inclined fluid layer .
we characterize the trajectories of the defects both by analogy to fluid turbulence
velocity distributions , diffusion , and power spectra and by properties dependent upon the topological characteristics of the defects : pair creation / annihilation and interactions . * ' '' '' |
the serpens cloud is a low mass star - forming cloud in the gould belt .
the cloud is known for its high star formation rate ( sfr ) and high surface density of young stellar objects ( ysos ) ( eiroa et al .
a 10 deg@xmath3 optical extinction ( a@xmath6 ) map made by cambresy ( 1999 ) originally defined the serpens cloud .
more recent studies treat the serpens cloud as two much smaller ( @xmath7 1.0 deg@xmath8 ) regions : serpens main ( centered on r.a .
@xmath9 , dec . @xmath10 ( j2000 ) ) and serpens south ( centered on r.a .
@xmath11 , dec . @xmath12 ( j2000 ) ) ( enoch et al . 2007 ;
harvey et al . 2007a ; gutermuth et al .
2008 ; eiroa et al .
2008 ; bontemps et al . 2010 ) .
serpens main is known mainly for its northernmost region , the serpens core , which has the highest yso surface density in the cambresy ( 1999 ) a@xmath6 map ( eiroa et al .
serpens south is part of the aquila rift molecular complex .
it was first studied in detail by gutermuth et al .
( 2008 ) , and has now been mapped at 70 - 500 @xmath13 m as part of the _ herschel _ gould belt survey ( andre et al .
2010 ; bontemps et al .
serpens main is the focus of this study ( see fig .
[ fig : mapped_reg ] ) .
the total molecular mass of the serpens core is uncertain by at least a factor of 5 .
some studies estimate @xmath7250 - 300 m@xmath14 ( mcmullin et al . 2000 ; olmi & testi 2002 ) while others find @xmath71450 m@xmath14 ( white et al .
these two results were based on c@xmath15o j=1 - 0 and c@xmath15o j=@xmath2 lines , respectively , so the large discrepancy may be due to the different gas properties traced by each c@xmath15o rotational transition ( eiroa et al .
the distance to serpens assumed by these studies may also be too low ( see section 1.2 ) .
clearly , the gas mass must be measured with better accuracy to determine the efficiency and history of star formation in the serpens cloud .
the best estimate for the distance to serpens main is @xmath16 pc , measured from vlbi trigonometric parallax of the yso , ec 95 ( dzib et al .
ec 95 is located at the center of the serpens core and is therefore almost certainly a member , so we adopt the dzib et al .
( 2010 ) 415 pc distance .
this is almost twice the previously accepted value of @xmath17 pc ( eiroa et al .
2008 ) , so care must be used in comparing physical properties derived using the lower distance with our results in this paper .
we mapped about 1.04 deg@xmath8 of serpens main in the co and @xmath1co j = @xmath2 emission lines .
our study complements existing survey data , from the _ spitzer _ c2d legacy ( evans et al .
2003 ; harvey et al . 2006
; harvey et al . 2007 ) and bolocam 1.1 mm continuum ( enoch et al .
2007 ) surveys .
other molecular tracer data exist for serpens main , such as n@xmath18h@xmath19 ( testi et al .
2000 ) , but these data are almost always limited to a sub - region of serpens main ( e.g. , the serpens core )
[ fig : mapped_reg ] shows our mapped region ( red polygon ) , the _ spitzer _
c2d regions with mips ( gray polygon ) and irac ( thin - black polygon ) , and the bolocam 1.1 mm region ( thick black polygon ) overlaid on the cambresy ( 1999 ) @xmath20 map of serpens main . the cambresy ( 1999 ) map suggests @xmath20 @xmath21 10 mag in serpens main , but the more recent @xmath20 map derived from c2d _ spitzer _ data ( enoch et al .
2007 ) shows @xmath20 @xmath22 25 mag .
continuum 1.1 mm emission reveals the locations of the coldest and densest dust .
n@xmath18h@xmath19 traces the highest density star forming gas ( testi et al .
2000 ) . from the _ spitzer _ c2d data , harvey et al . ( 2007 ) presented a high - confidence set " of 235 ysos associated with serpens main .
most ysos have masses @xmath23 m@xmath14 ( eiroa et al . 2008 ) , but there is at least one ( vv serpentis ) with mass @xmath24 m@xmath14 ( ripepi et al . 2007 ;
dzib et al .
enoch et al .
( 2007 ) identified 35 sub - mm sources in serpens main from the bolocam 1.1 mm survey .
we will compare the positions of these ysos and sub - mm sources with our co and @xmath1co data ( see section 5 ) .
we divide serpens main into three sub - regions with the following naming scheme : the _ serpens core _
( eiroa et al . 2008 ) , _ serpens g3-g6 _ ( cohen & kuhi 1979 ) , and _ vv serpentis _
( chavarria et al . 1988 ) .
these sub - regions are labeled _ 1 , 2 , _ and _ 3 _ , respectively , in fig .
[ fig : mapped_reg ] . other names previously used for the serpens core include serpens dark , serpens north , and cluster a ( harvey et al . 2007 ;
eiroa et al . 2008
; bontemps et al . 2010 ) ; cluster b for serpens g3-g6 ( harvey et al . 2007 ;
enoch et al .
2007 ) ; and cluster c for vv serpentis ( harvey et al . 2007 ;
eiroa et al .
2008 ) .
this study is a continuation of a molecular cloud mapping project with the arizona radio observatory .
previous papers in this series mapped the w51 region in co j = @xmath2 and @xmath1co j = @xmath2 ( bieging , peters , & kang 2010 ) and the w3 region in co j = @xmath2 and j = @xmath25 and @xmath1co j = @xmath2 ( bieging & peters 2011 ) .
for further details about the serpens cloud , we refer the reader to : harvey et al .
( 2007 ) ; enoch et al . ( 2007 ) ; gutermuth et al .
( 2008 ) ; eiroa et al .
( 2008 ) ; and bontemps et al .
( 2010 ) .
the goal of our study was to map the serpens core , serpens g3-g6 , and vv serpentis regions with high resolution in the j=@xmath2 rotational lines of @xmath0c@xmath26o and @xmath1c@xmath26o ( hereafter co and @xmath1co respectively ) .
co emission will be determined principally by cloud temperature and global turbulence .
in contrast , @xmath1co will generally be more useful as a measure of the column density .
section 2 describes our observations and data reduction techniques . in section 3
, we show our final brightness temperature image cubes for co and @xmath1co j=2 - 1 , and present associated velocity moment maps .
section 4 gives the website where our calibrated co and @xmath1co image cubes can be downloaded . in section 5
, we overlay the harvey et al .
( 2007 ) ysos , enoch et al .
( 2007 ) sub - mm sources , and testi et .
al ( 2000 ) n@xmath18h@xmath19cores on our co and @xmath1co maps .
section 6 is discussion , and section 7 is our summary .
the observations were made between 2008 november and 2010 june with the heinrich hertz submillimeter telescope ( hht ) on mt .
graham , az , at an elevation of 3200 m. the hht has a 10-m diameter paraboloidal dish and observes in the frequency range from 210 to 500 ghz . using prototype alma band 6 sideband separating mixers ( courtesy of the national radio astronomy observatory ) in a dual polarization receiver on the hht ( lauria et al .
2006 ) , we were able to take co ( 230.5 ghz ) and @xmath1co ( 220.4 ghz ) , j=@xmath2 , spectra simultaneously through the same telescope optics .
we used a filterbank spectrometer to record both the co and @xmath1co emission simultaneously .
each of the two j=@xmath2 lines was measured with 256 filters of 0.25 mhz bandwidth , giving a total spectral coverage of @xmath27 km s@xmath5 at a resolution of 0.33 km s@xmath5 .
co and @xmath1co j = @xmath2 emission in our mapped region of serpens spans a velocity range from @xmath28 to 20 km s@xmath5 ( lsr ) .
our spectra extend from @xmath29 to 30 km s@xmath5 ( lsr ) so we detect all gas in the cloud .
observations were made by the on - the - fly " ( otf ) method , where the telescope was scanned in a boustrophedonic pattern in right ascension at a scan rate of 10@xmath30/sec , with spectra sampled every 0.1 sec . in subsequent processing ,
the data were smoothed by 4 samples or 0.4 sec , corresponding to a telescope motion of 4@xmath30 per spectrum .
the beam size at the co j=@xmath2 observing frequency is 32@xmath30 ( fwhm ) so this choice of spatial sampling does not cause any significant loss of angular resolution in the scan direction .
a raster of lines with 10@xmath30 spacing in declination or @xmath31 of the beam fwhm was observed in this way . the telescope resolution is well - sampled in both coordinates with this choice of otf mapping parameters .
the field to be mapped was divided into 39 subfields , each @xmath32 in size , plus a small overlap .
each subfield was observed at least once in each of the two co isotopes .
total scanning time per subfield was therefore 2.0 hours , plus @xmath730% overhead for telescope slewing and measurement of reference spectra after every other row .
subfields were re - observed in whole or part if weather conditions resulted in excessively high noise levels .
the telescope pointing accuracy was checked at the start of each observing period by measuring the co j=@xmath2 emission from the red giant star v1111 oph , which has strong co emission lines from its circumstellar molecular envelope .
this object is a good pointing calibrator because the stellar emission is compact and centered on the well - known position of the star , which is only about 10@xmath33 from our serpens field .
a 5-point cross pattern centered on the star was measured and the integrated line intensities at the 5 positions were fitted to determine the pointing corrections in azimuth and elevation .
these corrections were generally @xmath34
. the pointing was checked and corrected about every 2 hours throughout each observing run , so the pointing errors during the otf mapping should be @xmath34 .
the intensity scale was determined with the standard method of kutner & ulich ( 1981 ) by comparing an ambient temperature load with on - sky spectra near the target field , to give the spectral line intensity on the @xmath35 scale .
the sideband rejections for the mixers were measured each time the receiver was tuned .
we observed a standard position , w51d , to place all the otf data on a consistent main - beam brightness temperature scale , in the manner described in bieging et al .
( 2010 ) .
all of the otf data were processed in the _ class _ software package of the grenoble astrophysics group .
a linear baseline was removed from each spectrum and the data for each spectral channel were convolved to a square grid in r.a . and dec .
, with a 10@xmath30 grid spacing in both coordinates .
the gridding algorithm uses a circular gaussian weighting function with a fwhm of 0.3@xmath36(beam fwhm ) , to convolve the otf - sampled spectra onto the regular grid points .
the gridding thereby increased the effective angular resolution by a factor of @xmath37 .
the resulting data cubes were transferred to the _ miriad _ software format ( sault , teuben , & wright 1995 ) for subsequent processing and analysis . to facilitate the comparison of co and @xmath1co lines ( for example , to calculate line ratios at each pixel ) , we used the _ miriad _ task _ regrid _ to resample the data cubes on the velocity axis by 3rd order interpolation .
we chose to sample velocities at 0.15 km s@xmath5 intervals from @xmath29 to @xmath38 km s@xmath5 ( lsr ) , so that the original velocity resolution was oversampled by a factor of 2.2 . to match the angular resolutions of the co and @xmath1co j=@xmath2 maps , which differ by @xmath39 because of the difference in line rest frequencies , we convolved the images with gaussians chosen to give identical resolutions for the two isotopologues .
the effective resolution of the final maps was 38@xmath30 ( fwhm ) for both the co and @xmath1co images .
this additional spatial smoothing also reduced the surface brightness noise level , at only a small cost in angular resolution .
we calculated the rms noise in an emission - free velocity range for each map pixel .
[ fig : rms_map ] shows the resulting distribution of the noise . the co j=@xmath2 image _ ( left ) _ has relatively uniform noise over the whole field , with a mean value of @xmath40 k ( main beam brightness ) per pixel and per velocity channel , with variations of @xmath41 about this value over most of the map . for the entire @xmath1co image _ ( right ) _ the mean @xmath42 k with comparable levels of variation due to differences in weather ( i.e. , sky noise ) and source elevation when the individual subfields were observed .
in fig . [ fig : mean_spectra ] we show the spatially averaged co and @xmath1co spectra for the serpens core ( region 1 ) , serpens g3-g6 ( region 2 ) , and vv serpentis ( region 3 ) in fig . [
fig : mapped_reg ] .
the radial velocity of the emission ranges from about @xmath43 to 18 km s@xmath5 for co and from 2 to 13 km s@xmath5 for @xmath1co .
the peak brightness temperature for @xmath1co occurs at 8 km s@xmath5 in all three regions , consistent with previous studies ( duarte - cabral et al .
2010 ; graves et al .
2010 ) which concluded that this is the velocity of the bulk of the gas in serpens .
we adopt 8 km s@xmath5 as the systemic velocity of serpens ( indicated by the vertical line in fig .
[ fig : mean_spectra ] ) .
the @xmath1co spectra for all three regions are nearly gaussian in shape .
the co profiles , in contrast , are markedly asymmetric and show a dip or inflection at 8 km s@xmath5 , the velocity of the @xmath1co peak emission .
this comparison strongly suggests that the co lines are self - absorbed over at least some of the brightest emission regions , possibly as a result of colder gas which is optically thick in co and is located on the near side of a warmer central part of the cloud .
[ fig : co_channel ] shows representative velocity slices from our co image cube , averaging over 0.6 km s@xmath5 and stepping by 1.2 km s@xmath5 . at velocities of about 4.8 and 10.8 km
s@xmath5 the emission appears elongated and filamentary .
[ fig:13co_channel ] shows a similar set of velocity slices from our @xmath1co image cube , averaging over 0.6 km s@xmath5 , at the same velocities as in fig .
[ fig : co_channel ] for the brightest co emission ( 4.4 to 10.4 km s@xmath5 ) . outside this velocity range ,
@xmath1co is mostly below our detection limit .
the @xmath1co j=@xmath2 emission , which is generally optically thin , should be a good tracer of molecular column density in the cloud .
[ fig : peak_bt ] shows maps of maximum brightness temperature ( @xmath44 ) in co and @xmath1co , over the full range of emission , -11 to 30 km s@xmath5 for co and -5 to 20 km s@xmath5 for @xmath1co .
the co @xmath44 distribution is very clumpy with temperatures of 5 to 10 k in serpens g3-g6 and parts of vv serpentis , and temperatures of 15 to 25 k in the serpens core . when co is not self - absorbed or depleted , co @xmath44 approaches the value of the kinetic temperature of the gas ( @xmath45 ) .
in contrast , @xmath1co @xmath44 is much more uniform .
we present integrated co and @xmath1co brightness temperature ( @xmath46 ) maps in fig .
[ fig : integrated_bt ] , displayed in units of k km s@xmath5 .
we only consider velocities associated with serpens by summing over -1 to 18 km s@xmath5 for co , and 2 to 13 km s@xmath5 for @xmath1co .
we use a 3 rms cutoff to ensure that detected emission is real .
@xmath1co is generally optically thin and its spectrum well - behaved , so the @xmath1co integrated @xmath46 map is nearly identical to the @xmath1co @xmath44 map .
@xmath1co integrated @xmath46 should trace the highest column density gas , as long as @xmath1co is optically thin . compared to @xmath44 for co ,
co integrated @xmath46 retains its clumpy structure in the serpens g3-g6 region , but is smoothly distributed otherwise . we show centroid velocity maps ( first - moment ) for co and @xmath1co in fig .
[ fig : vel_centroid ] , considering the same range of velocities as we used to make the integrated @xmath46 maps .
note that even where co is self - absorbed , the first moment is the mean velocity of the entire line profile and not ( necessarily ) the velocity of the highest peak bracketing the self - absorption .
the color palette extends @xmath47 3 km s@xmath5 on either side of 8 km s@xmath5 , the systemic lsr velocity of serpens .
red colors are redshifted with respect to the systemic velocity .
the centroid velocity of co in the serpens core and @xmath1co in all regions is mostly at the systemic velocity of serpens and spatially uniform .
this uniformity contrasts with the centroid velocity of co southward of the serpens core , which is not spatially uniform but appears in filamentary structures with lsr velocities @xmath47 3 km s@xmath5 of the systemic velocity of serpens .
this difference in apparent spatial distribution is most likely due to a change in the velocity structure and/or the opacity of the colder absorbing co outside the serpens core .
there are two kinematically distinct areas in the serpens core , seen as red and blue components in co centroid velocity separated by about 3 km s@xmath5 .
it has been suggested that these are two interacting sub - clouds ( testi et al .
2000 ; eiroa et al .
2008 ; duarte - cabral et al . 2010 ) .
[ fig : vel_width ] shows velocity dispersion ( @xmath48 ) ( second - moment ) maps for co and @xmath1co , in units of km s@xmath5 .
we apply a 3 rms cutoff @xmath49 k for both isotopes .
the velocity range for computing the co @xmath48 map is -1 to + 16 km s@xmath5 ( which avoids the high velocity gas component at r.a .
@xmath50 , dec .
@xmath51 ) , and 4 to 12 km s@xmath5 for the @xmath1co @xmath48 map .
the maximum co @xmath48 is 4.2 km s@xmath5 , and that of @xmath1co is 2.0 km s@xmath5 , while the average co @xmath48 is 2 km s@xmath5 , and that of @xmath1co is 0.8 km s@xmath5 .
large @xmath48 reveals extended emission - line wings , which can be caused by large scale gas motions or turbulence due to protostellar outflows .
the @xmath1co map is relatively noisy ( i.e. , pixelated ) for small ( @xmath52 km s@xmath5 ) @xmath48 , which occurs away from regions of high column density . unlike @xmath1co @xmath44 , integrated @xmath46 , and large co @xmath48 ( @xmath53 km s@xmath5 )
, large @xmath1co @xmath48 ( @xmath54 km s@xmath5 ) occurs in small ( 0.01 x 0.01 deg@xmath8 ) regions .
there is one such region in the serpens core and three to five in serpens g3-g6 .
note that we exclude a @xmath55 high velocity ( 18 - 20 km s@xmath5 ) gas component ( r.a .
@xmath50 , dec .
@xmath51 ) from our moment maps .
we assume that this feature is not associated with serpens main . following bieging et al .
( 2011 ) , fig .
[ fig : ratio_map ] shows the ratio ( @xmath56 ) of co to @xmath1co line intensities ( @xmath46 ) , averaged over 0.6 km s@xmath5 and spaced 0.6 km s@xmath5 , as a function of velocity . for an assumed @xmath0co/@xmath1co abundance ratio of 50 , the color palette indicates where @xmath1co reaches optical depth @xmath57 for @xmath58 ( gray regions ) , @xmath59 for @xmath60 ( colored regions ) , and co is self - absorbed for @xmath61 ( white regions ) .
there is a spatial gradient in optically thick @xmath1co across serpens , appearing in vv serpentis at low velocities and shifting north - easterly until @xmath1co is optically thick in the serpens core at higher velocities .
co is self - absorbed and @xmath1co is optically thick in the serpens core . the spatial and velocity structure of the co line ratios imply that a detailed 3-dimensional radiative transfer model of the cloud will be necessary for proper interpretation of these data .
the calibrated brightness temperature image cubes of co and @xmath1co j=2 - 1 can be downloaded as fits files from the online version of this paper through the astrophysical journal supplement series .
the evolutionary stage of a yso ( after protostar formation but before reaching the main sequence ) can be inferred from its observed sed ( slope of log@xmath64 f@xmath65 vs. log@xmath64 , @xmath66 ) ( lada 1987 ; andre & montmerle 1994 ; greene et al .
1994 ) and/or comparing its sed to theoretical models ( whitney et al . 2003
; robitaille et al .
these evolutionary stages are ( in order of youngest to oldest ) : class i , flat , class ii , and class iii ( lada 1987 ; andre & montmerle 1994 ; greene et al .
harvey et al . ( 2007 ) classified their high - confidence set of 235 serpens ysos into the above four stages , 41 class i ( 17 % ) , 25 flat ( 11 % ) , 130 class ii ( 55 % ) , and 39 class iii ( 17 % ) .
of the 15 ysos with the coldest seds " ( 70 to 24 @xmath67 flux ratio @xmath68 , harvey et al .
2007 ) , 10 are class i. these 10 ( which we refer to as cold class i ) should be the youngest class i ysos as they are the most embedded or obscured . note that harvey et al .
( 2007 ) state that there are 39 class i and 132 class ii ysos , while the online supplementary table from harvey et al .
( 2007 ) shows 41 class i and 130 class ii .
we use the data from the harvey et al .
( 2007 ) online supplementary table . in fig .
[ fig : yso_olay ] , we compare our @xmath1co integrated intensity map with locations of the harvey et al .
( 2007 ) ysos : cold " class i ( magenta @xmath69 s ) , all other class i ( magenta @xmath70 s ) , flat ( yellow @xmath70 s ) , class ii ( white @xmath69 s ) , and class iii ( white @xmath70 s ) .
the distribution of class i and flat ysos follows regions of large ( @xmath71 k km s@xmath5 ) @xmath1co integrated line intensity
. there are three class i ysos ( near r.a .
18@xmath7228@xmath7345@xmath74 , dec .
@xmath75 ) that do not coincide with large @xmath1co integrated line intensity ( see section 5.3 ) . in the serpens core ,
all cold " class i ysos are in locations of peak ( @xmath76 k km s@xmath5 ) @xmath1co integrated line intensity .
class ii and iii ysos appear more widely distributed across serpens main .
[ fig : serpens_core ] shows the @xmath1co velocity centroid ( 1st moment ) for the serpens core computed over line center ( 6 to 10.5 km s@xmath5 ) in red - green - blue palette , with positions of n@xmath18h@xmath19 j=@xmath77 cores ( cyan @xmath69 s with diameter 1 ) from testi et al .
( 2000 ) and ysos ( symbols as in fig .
@xmath78 ) from harvey et al . 2007 .
the n@xmath18h@xmath19 cores and class i and flat ysos form a cluster that is strikingly elongated , with a major axis nearly parallel to the boundary separating the red and blue - shifted gas in the serpens core .
the four n@xmath18h@xmath19 cores , which should mark the coldest , densest pre - stellar locations in the cloud , also coincide closely with the youngest ysos .
class ii and iii ysos are less clustered and more symmetrically distributed about this axis .
this distribution raises the question of whether the younger ( class i and flat ) or older ( class ii and iii ) ysos , or both , formed along this boundary . to examine these possibilities ,
we define two reference positions for serpens core ysos : the unweighted centroid r.a . ,
dec . of cold class
i ysos ( _ cold centroid _ ,
@xmath79 , dec . @xmath80 ) and that of flat ysos ( _ flat centroid _ , r.a .
@xmath81 , dec .
@xmath82 ) . fig .
[ fig : serpens_core ] shows these two positions as black @xmath83s .
the flat centroid lies within a tight grouping of flat ysos @xmath842 southeast of the cold centroid .
we note that these two centroid positions lie near the centers of the two sub - clusters " described by duarte - cabral et al .
( 2010 ) , based on their observations of the very optically thin c@xmath85o and c@xmath15o isotopologues .
nw sub - cluster " encompasses the position of our _ cold centroid _ ,
consistent with a very young age for the ysos contained in this sub - cluster .
in contrast , duarte - cabral et al . (
2010 ) find that the se sub - cluster " displays greater kinematic complexity and a wider distribution of yso ages .
this se sub - cluster contains our _ flat centroid _ position , consistent with the observed clustering of the flat - sed sources having somewhat greater ages than the cold class i sources . in figures [ fig : yso_hist_sep ] and [ fig : yso_hist_pa ] , we plot histograms of yso projected separation and position angle ( pa ) relative to the cold ( solid black ) and flat ( dashed blue ) centroids , separated by yso class .
( here class i " includes cold class i " ysos . )
the sample contains only ysos within the serpens core .
this restriction gives 23 class i , 12 flat , 23 class ii , and 7 class iii objects .
the histograms for the youngest classes ( i and flat ) in fig .
13 show that these ysos cluster within 0.5 pc of either centroid position , with small offsets ( @xmath70.2 pc ) as expected from the definitions of the 2 centroids .
the distribution of pas for the youngest classes ( fig .
14 ) clearly shows that the clustering is oriented along a nw - se ( @xmath86 ) axis , an orientation which is consistent with the location of the sub - clusters " discussed by duarte - cabral et al .
in contrast , the older ysos , classes ii and iii , show a wider distribution relative to either of the centroids ( fig .
13 ) , with the class iii ysos lying almost entirely at separations @xmath87 pc , while the class ii ysos are broadly peaked out to @xmath71 pc .
both the class ii and iii objects are uniformly distributed in pa , with no preference for the nw - se axis of the younger yso classes .
the spatial distrobutions of the various yso classes suggest that the most recent star formation in the serpens core , i.e. , within the past few@xmath88 years , has occurred almost entirely within a volume of space about 1 pc in length and elongated in a nw - se direction as projected on the sky .
the wider distribution of the class iii objects could be a result of their having formed over a more extended volume in an earlier episode of star formation , as suggested by duarte - cabral et al .
the class ii ysos , however , show a distribution that is clustered about the reference centroid positions but less tightly than the class i or flat ysos
. the progression from a more concentrated to a more dispersed distribution with age could also result from a diffusion of ysos outward away from the serpens core over the past few million years , due to random motions acquired from the internal velocity dispersion of the parent gas clouds .
davis et al . ( 1999 ) presented co j=2 - 1 maps of the serpens core region ( @xmath89 ) with 23 resolution , and found evidence in the line wings ( 14 to 22 km s@xmath5 and @xmath90 to 4 km s@xmath5 ) for a burst of outflows " , produced by the cluster of ysos in the core .
our data have higher sensitivity but lower angular resolution ( 38 ) than davis et al .
( 1999 ) ; a detailed comparison with their line wing maps shows excellent agreement for both the red- and blue - shifted components , including the broad red - shifted wing associated with the herbig - haro object hh106 ( located @xmath91 west of the core cluster ) .
more recently , graves et al .
( 2010 ) have mapped the j=3 - 2 lines of co , @xmath1co , and c@xmath15o over the serpens core with the harp heterodyne receiver array on the jcmt , with effective resolutions of 17 to 20 .
their co and @xmath1co integrated intensity maps ( their fig .
2 ) agree well with ours ( fig .
7 ) , allowing for the different transitions and angular resolution . at the position of hh106 ( 18@xmath7229@xmath7318@xmath74 , @xmath92 ) ,
the co j=2 - 1 line shows the same broad red - shifted wing as their j=3 - 2 line ( their fig .
5 ) , but the lower excitation j=2 - 1 transition has a more pronounced narrow self - absorption feature at 8 km s@xmath5 , the systemic velocity of the cloud .
our line ratio channel maps ( fig .
7 ) show that near the systemic velocity , much of the serpens core cloud is optically thick in @xmath1co j=2 - 1 , and the co line is likely to be self - absorbed .
graves et al .
( 2010 ) also find from the j=3 - 2 line ratios that the @xmath1co line is optically thick with @xmath93 having values up to @xmath77 toward the densest submm continuum cores .
duarte - cabral et al . (
2010 ) presented maps of the c@xmath85o and c@xmath15o j=2 - 1 and 1 - 0 transitions made with the iram 30 m telescope , covering only the central [email protected] of the serpens core .
these optically thin transitions should be less distorted by opacity effects than the @xmath1co or co lines , revealing more accurately the distribution and kinematics of the highest column density regions . duarte - cabral et al .
( 2010 ) found that the sub - clusters have different kinematic properties .
the nw center has a single , nearly gaussian line profile , while the se center has 2 blended components .
they interpret the profiles as a superposition of 2 clouds with different velocities and argue that these are colliding in the area containing the se sub - cluster . in a second paper , duarte - cabral et al .
( 2011 ) show that a smoothed particle hydrodynamics ( sph ) 3d simulation of two colliding gas cylinders , one rotated @xmath7 45@xmath33 relative to the other , can reproduce many of the observational properties of the serpens core .
especially the onset of star formation in the zone where the clouds collide .
their sph simulation includes gas hydrodynamics , self - gravity , turbulence , and sink particles .
given the large differences in optical depths of the c@xmath85o and c@xmath15o lines compared with the co and @xmath1co transitions we present here , it is difficult to make direct comparisons .
we note however , that duarte - cabral et al .
( 2010 ) find a velocity gradient in the c@xmath15o j=1 - 0 map with velocity increasing from the se to nw sub - clusters , that is consistent with that in our @xmath1co velocity centroid map ( fig .
our data show a gradient over a much larger region ( @xmath710 ) than that covered by duarte - cabral et al .
( 2010 , 2011 ) .
their sph simulation may therefore underestimate the physical size of the colliding clouds . in fig .
[ fig : bolocam_olay ] , we overlay contours of 1.1 mm continuum emission ( enoch et al . 2007 ) on our @xmath1co integrated @xmath46 map ( convolved to 90 resolution to match that of the published bolocam map ) .
contour levels are at 5 , 10 , 20 , 40 , 80 , and 160 times the rms noise @xmath94 mjy beam@xmath5 .
the 1.1 mm emission coincides with high @xmath1co integrated @xmath46 , except for a filamentary feature in serpens g3-g6 centered at r.a .
@xmath95 , dec .
@xmath75 called the starless cores region " ( enoch et al .
i ysos and two class ii ysos are located in the starless cores region ( see fig .
[ fig : yso_olay ] ) .
we detect only a faint ( @xmath96 k km s@xmath5 ) local maximum in @xmath1co integrated @xmath46 at r.a .
@xmath97 , dec . @xmath98 .
submm maps of the serpens core have been presented by davis et al .
( 1999 ) , made with the scuba bolometer array at wavelengths of 850 and 450 .
these images have considerably better resolution ( 14 and 8 respectively ) but like the bolocam 1.1 mm map , also suffer from spatial filtering so that mainly the bright compact sources are detected .
four of these ( smm 1 , 5 , 9 , and 10 ) lie within the nw sub - cluster and lie on the secondary maximum of the integrated @xmath1co intensity ( fig . 7 ,
right ) @xmath73 nw of the brightest @xmath1co peak .
the other 6 submm compact sources are associated with the brightest peak in fig . 7 ( right ) , and with the se sub - cluster of duarte - cabral et al .
this positional agreement implies that , as expected , the current most active star formation coincides with the region of greatest gas column density .
serpens main has yet to be extensively studied using data from the _ herschel _ gould belt survey , which spans far - ir to sub - mm wavelengths .
such data could reveal even younger prestellar cores ( class 0 ) in serpens main .
for example , goicoechea et al . (
2012 ) presented a submm spectrum for a single serpens core class 0 yso .
hundreds of class 0 ysos have been identified in the nearby aquila rift molecular complex using _ herschel _ data ( andre et al .
2010 ; bontemps et al . 2010 ) .
the spatial distribution of co , as seen in the peak and integrated brightness temperature maps ( figs . 6 & 7 ) as well as the individual channel maps ( fig .
4 ) gives the impression of a highly disturbed gas cloud .
there are shell - like structures and holes especially in the serpens core region .
elsewhere the co has a flocculent appearance with many small clumps ( @xmath70.1 ) all having approximately the same peak brightness temperature .
overall , there is a gradient of increasing peak and integrated brightness temperatures from south to north over the 2 declination extent of our maps . although co emission is detected in every pixel , there are large contrasts in some regions .
the serpens core , at the northern end of the cloud , stands out sharply against a lower @xmath99 background , as a result of the high concentration of ysos which energize the gas by radiation heating of the associated dust , and by injecting kinetic energy via numerous bipolar outflows .
the class i and flat ysos in the serpens core form a cluster that is nearly parallel to the intersection of the red and blue - shifted gas components and is strikingly linear - shaped ( see fig .
[ fig : serpens_core ] ) , similar to the corresponding distribution of sub - mm cores ( enoch et al . 2007 ; duarte - cabral et al . 2010 ) .
the velocity gradient might result from a collective action of outflows in the center of the serpens core , and the linear - shaped cluster might indicate a preferred orientation for proto - stellar jets .
duarte - cabral et al . (
2011 ) show that a cloud - cloud collision also explains the velocity gradient and linear - shaped cluster .
however , their sph simulation should be treated with caution until the result is independently verified using an adaptive mesh refinement ( amr ) code , such as athena or flash , to calculate accurately the effects of shocks in the collision . from the observed @xmath1co velocity dispersion ( fig .
[ fig : vel_width ] ) and present yso distribution ( fig .
[ fig : yso_hist_sep ] ) , we find that class i , flat , and class ii ysos in the serpens core had sufficient drift velocities and lifetimes to reach their present locations , even if they all formed in the same limited volume . most class i and flat ysos are within @xmath840.5 projected pc of both the flat and cold centroids , while most class ii ysos are within @xmath841.0 projected pc ( see fig . [
fig : yso_hist_sep ] ) . assuming a mean inclination angle of 45@xmath33 to the line of sight
, class i and flat , and class ii ysos would need to drift @xmath840.7 pc and 1.4 pc , respectively .
class ii ysos have lifetimes @xmath84 2 myr , which sets class i and flat yso lifetimes at @xmath840.5 and 1.0 myr , respectively ( evans et al .
assuming drift velocity ( @xmath100 ) is constant since yso formation , class i , flat , and class ii ysos would need minimum @xmath100 @xmath842.0 , 1.0 , and 1.0 km s@xmath5 , respectively . because co traces bulk gas motion , we can assume that @xmath100 @xmath84 @xmath101 ( i.e. , protostars inherit the velocity dispersion of accreted gas ) , where @xmath101 is the spatial average of @xmath102 .
we calculate @xmath101 over r.a . @xmath103
@xmath104 and dec .
@xmath106 , the smallest region containing the yso sample from figs .
@xmath101 = 2.0 km s@xmath5 and rms @xmath107 km s@xmath5 , giving @xmath100 @xmath108 km s@xmath5 .
@xmath100 exceeds or equals what is necessary for nearly all class i , flat , and class ii ysos to have formed in a small volume at either the flat or cold centroid positions .
since we do not detect co and @xmath1co in the starless cores " region , co is most likely depleted there .
average co @xmath44 from fig .
[ fig : peak_bt ] is @xmath8410 - 20 k , so @xmath45@xmath8410 - 20 k ( assuming no co self - absorption or depletion ) . at 10 - 20 k co becomes depleted for number densities @xmath109 @xmath110 ( goldsmith 2001 ) .
the lowest ( inferred ) gas number density that bolocam can detect is @xmath84@xmath111 @xmath110 ( enoch et al .
co is unlikely to be depleted in the serpens core and serpens g3-g6 regions because the temperature of the surrounding gas is too high , so co will not freeze out onto dust grains , despite sufficiently high number densities .
there are in fact five ysos in the starless cores " region ( i.e. , not truly starless ) , but heating may be insufficient ( co and @xmath1co peak t@xmath112 is about 3 and 1.5 k , respectively ) to prevent co from freezing out onto gains .
we mapped the j=@xmath2 rotational lines of co and @xmath1co over 3900 square arcmin of the serpens core , serpens g3-g6 , and vv serpentis regions of the serpens main cloud ( 1.04 deg@xmath8 in total ) with high spatial ( 38 ) and spectral ( 0.3 km s@xmath5 ) resolution .
our final data are calibrated @xmath46 image cubes for co and @xmath1co ( available online in fits format ) .
the _ spitzer _
c2d legacy survey ( evans et al .
2003 ; harvey et al . 2007 ) and the bolocam 1.1 mm continuum survey ( enoch et al .
2007 ) are the only other studies that have mapped the entirety of the serpens main molecular cloud .
these latter two surveys were limited to measuring dust temperature and column density ; our co and @xmath1co data measure gas kinematics as well as physical properties . at the systemic velocity of serpens main ( 8 km s@xmath5 ) ,
co is self - absorbed and @xmath1co is optically thick in the serpens core .
the gas traced by co in serpens g3-g6 and vv serpentis appears in filamentary structures having lsr velocities between 6 and 8 km s@xmath5 .
the spatial and velocity structure of the co line ratios implies that a detailed 3-dimensional radiative transfer model of the cloud would be necessary for a complete interpretation of our spectral data .
it is likely that the observed class i , flat , and class ii ysos in the serpens core formed in a @xmath71 pc volume between the flat and/or cold centroid positions .
the distributions of projected separations and pas suggest a co - spatial formation site ( figs . [
fig : yso_hist_sep]-[fig : yso_hist_pa ] ) .
our measured @xmath1co velocity dispersion of @xmath84@xmath113 km s@xmath5 implies that the serpens core ysos could have formed within a small volume and then diffused away to their observed spatial distributions on a timescale of @xmath114 yr . the blue- and red - shifted regions in the @xmath1co velocity first moment ( fig .
[ fig : serpens_core ] ) could be the result of a collective action of outflows , such as a preferred protostellar jet orientation .
an alternative explanation is a cloud - cloud collision between the blue - shifted gas and foreground gas along the line of sight ( duarte - cabral et al .
2010 ; duarte - cabral et al .
2011 ) .
the starless cores " region is likely to be the site of further star formation in serpens .
the detection of 1.1 mm cold dust emission but no co or @xmath1co emission in that region , suggests that co is largely depleted due to high density and low termperatures in this part of the serpens cloud . in a future paper
we will employ a grid of statistical equilibrium / radiative transfer models for co line emission , incorporating co excitation and photodissociation , to derive the distribution of total hydrogen ( h@xmath18 and hi ) column densities and masses of the serpens core , serpens g3-g6 , and vv serpentis regions .
this analysis should resolve the disagreement between previous mass estimates noted by eiroa et al .
this material is based upon work supported by the national science foundation graduate research fellowship under grant no .
dge 1106400 .
this work was supported in part by national science foundation grant ast-0708131 to the university of arizona , and by the nasa space grant program .
we thank the anonymous referee for suggestions that strengthened the discussion and conclusion sections in this paper .
we thank steve stahler , chris mckee , yancy shirley , neil evans , and kevin hardegree - ullman for helpful discussions .
we also thank dr .
a. r. kerr of the national radio astronomy observatory for providing the prototype alma band 6 mixers used in this work .
we are grateful to butler burton and greg schwarz for a prompt submission process and storing our data online with the astrophysical journal supplemental series .
the heinrich hertz submillimeter telescope is operated by the arizona radio observatory , which is part of steward observatory at the university of arizona .
andre , a. , & montmerle t. 1994 , apj , 420 ,
837 andre , ph . , menshchikov , a. , bontemps , s. , et al .
2010 , a&a , 518 , l102 bieging , j. h. , peters , w. l. , & kang , m. 2010 , apjs , 191 , 232 bieging , j. h. , & peters , w. l. 2011 , apjs , 196 , 18 bontemps , s. , andre , ph . , konyves , v. , et al .
2010 , a&a , 518 , l85 cambresy , l. 1999 , a&a , 345 , 965 chavarria , k. , lara , e. de , finkenzeller , u. , mendoza , e. e. , & ocegueda , j. 1988 , a&a , 197 , 151 cohen , m. , & kuhi , l. v. 1979 , apjs , 41 , 743 davis , c.j . ,
matthews , h.e .
, ray , t.p . , dent ,
w.r.f , & richer , j.s .
1999 mnras , 309 , 141 duarte - cabral , a. , fuller , g. a. , peretto , n. , et al .
2010 , a&a , 519 , a27 duarte - cabral , a. , dobbs , c. l. , peretto , n. , & fuller , g. a. 2011 , a&a , 528 , a50 dzib , a. , loinard , l. , mioduszewski , a. j. , et al .
2010 , apj , 718 , 610 eiroa , c. , djupvik , a. a. , & casali , m. m. 2008 , in _ handbook of star forming regions ii : the southern sky _ , ed . b. reipurth , ( asp monograph publications ) , 693 enoch , m. l. , glenn , j. , evans , n. j. , et al .
2007 , 666 , 982 evans , n. j. , allen , l. e. , blake , g. a. , et al .
2003 , pasp , 115 , 965 evans , n. j. , dunham , m. m. , jorgensen , j. k. , et al .
2009 , apj , 181 , 321 goicoechea , j. r. , 2012 , a&a , 548 , 17 pp goldsmith , p. f. 2001 , apj , 557 , 736 graves , s. f. , richer , j. s. , buckle , j. v. , et al .
2010 , mnras , 409 , 1412 greene , t. p. , wilking , b. a. , andre , p. , young , e. t. , & lada , c. t. 1994 , apj , 434 , 614 gutermuth , r. a. , bourke , t. l. , allen , l. e. , et al .
2008 , apj , 673 , l151 harvey , p. m. , chapman , n. , lai , s. , et al .
2006 , apj , 644 , 307 harvey , p. m. , merin , b. , huard , t. l. , et al .
2007 , apj , 663 , 1149 kutner , m. l. , & ulich , b. l. 1981 , apj , 250 , 341 lada , c. j. 1987 , in _ star forming regions _ , eds . m. peimbert & j. jugaku ( dordrecht : reidel ) , 1 lauria , e.f . ,
kerr , a.r . ,
reiland , g. , et al .
2006 , alma memo 553 , ( nrao : charlottesville , va ) mcmullin , j. p. , mundy , l. g. , blake , g. a. , et al . 2000 , apj , 536 , 845 olmi , l. , & testi , l. 2002 , a&a , 392 , 1053 ripepi , v. , bernabei , s. , marconi , m. , et al .
2007 , a&a , 462 , 1023 sault , r. j. , teuben , p. j. , & wright , m. c. h. 1995 , in _ astronomical data analysis software and systems iv _
r. shaw , h. e. payne , & j. j. e. hayes ( asp conf .
77 ) , p. 433 testi , l. , sargent , a.i .
, olmi , l , & onello , j.s .
2000 , apj , 540 , l53 white , g.j . ,
casali , m.m . , & eiroa , c. 1995 , a&a , 298 , 594 | we mapped @xmath0co and @xmath1co j = @xmath2 emission over 1.04 deg@xmath3 of the serpens molecular cloud with @xmath4 spatial and 0.3 km s@xmath5 spectral resolution using the arizona radio observatory heinrich hertz submillimeter telescope .
our maps resolve kinematic properties for the entire serpens cloud .
we also compare our velocity moment maps with known positions of young stellar objects ( ysos ) and 1.1 mm continuum emission .
we find that @xmath0co is self - absorbed and @xmath1co is optically thick in the serpens core . outside of the serpens core
, gas appears in filamentary structures having lsr velocities which are blue - shifted by up to 2 km s@xmath5 relative to the 8 km s@xmath5 systemic velocity of the serpens cloud .
we show that the known class i , flat , and class ii ysos in the serpens core most likely formed at the same spatial location and have since drifted apart .
the spatial and velocity structure of the @xmath0co line ratios implies that a detailed 3-dimensional radiative transfer model of the cloud will be necessary for full interpretation of our spectral data .
the starless cores " region of the cloud is likely to be the next site of star formation in serpens . |
low mass x - ray binaries ( lmxbs ) are considered to be the progenitor of the millisecond pulsars detected in the radio band ( see @xcite for an extensive review ) .
it is widely believed that the weakly magnetized , slowly rotating neutron star is gradually spun - up through the transfer of angular momentum carried by the matter from an accretion disk @xcite .
millisecond - time - scale variations have been observed in many lmxb systems , including kilohertz quasi - periodic oscillation ( khz qpo , see * ? ? ?
* ) and burst oscillation ( see @xcite for a review ) ; however , the attempts to measure coherent millisecond pulsation from an lmxb , which provides a direct evolutionary link between radio millisecond pulsar and lmxb , were unsuccessful until the discovery of the first accretion - powered millisecond pulsar sax 1808.4 - 3658 with a spin period @xmath5 ms @xcite , in 1998 .
to date , coherent millisecond pulsations have been detected in ten lmxbs with spin periods ranging from 1.67 ms to 5.5 ms @xcite .
the fourth known accretion - powered millisecond pulsar , xte j1807 - 294 , was discovered on february 21 , 2003 while the proportional counter array ( pca ) onboard the _ rossi x - ray timing explorer _ ( _ rxte _ ) monitored the galactic - center region @xcite .
its preliminary spin frequency of 190.625 hz @xcite and orbital period of @xmath6 min @xcite were immediately revealed right after its discovery .
follow - up observations were made with many x - ray observatories .
@xcite determined the source location , r.a.=@xmath7 dec=@xmath8 ( equinox 2000 , @xmath9 error ) by using _
chandra _ observation .
@xcite , @xcite and @xcite reported more precise orbital parameters and pulsation period using _ xmm - newton _ observations .
however , @xcite found wide swings in the apparent spin change rate vs. accretion rate , which were interpreted as the hot spot moving on the neutron star rather than from accretion torque . the wide - band ( 0.5 to 200 kev ) spectrum obtained by the combination of simultaneous observations from _ xmm - newton _ , _ rxte _ and _ integral _
is well described by an absorbed disk blackbody plus thermal comptonization model @xcite .
_ rxte _ has processed extensive monitoring of xte j1807 - 294 since february 27 , 2003 with a time span of more than 150d until the source went into its quiescent state .
the light curve exhibited an exponential - like decay with a time - scale of @xmath10d at the beginning of the outburst followed by a much slower decay with a time constant of @xmath11d @xcite .
several broad soft flares lasting for hours to days were seen during the flux decay .
the pulsed fraction increased and the pulse profile became more sinusoidal during the flares @xcite .
twin khz qpos were also detected in xte j1807 - 294 by _
rxte _ @xcite . owing to their fast rotation ,
the related parameters of the accretion - powered millisecond pulsars , such as orbital and spin parameters , can be precisely measured during their outbursts lasting for tens to hundreds of days .
unfortunately , the orbital and spin parameters derived by @xcite , @xcite and @xcite for xte j1807 - 294 form _
xmm - newton _ observations were based on the presumption that the orbital period is fixed to the preliminary value proposed by @xcite because the time base line is too short to constrain the orbital period . on the other hand ,
the @xmath0150d s _ rxte _ observations allow further binary parameter refinement , that is essential for further investigation . in this paper , we report the pulsation analyses of xte j1807 - 294 using archived _ rxte _ data ( section [ dar ] ) . due to the complexity of pulse phase variation , a trend removal technique is employed to reveal precise orbital parameters for the system .
pulse phases ( binary barycenter - corrected ) are found to exhibit anomalous negative shifts during the soft flares in all energy bands between 2 - 20 kev .
the evolution of the non - flare pulse phases shows that the neutron star was spun - up during the outburst .
significant @xmath12 @xmath4 soft lags for the non - flare pulses is detected at energies from 2 to 20 kev .
analysis results , including possible implications for coincident soft flares and anomalous negative phase shifts are discussed in section [ dis ] .
extensive follow - up observations for xte j1807 - 294 were conducted by _
rxte _ soon after the source was discovered until it went into its quiescent state .
we analyzed the available archived _ rxte _ data observed from february 27 to july 29 , 2003 .
the data used to study the timing properties of xte j1807 - 294 were collected by _ rxte _
pca @xcite , which consists of five gas - filled proportional counter units ( pcus ) with the total collecting area @xmath06500 @xmath13 and sensitive in the photon energy range 2 to 60 kev . the pca data of all the observations were collected in the goodxenon mode with a time resolution of @xmath01 @xmath14s .
in addition , to compare the pulsation behavior and light curve ( see section [ ppfc ] ) , the data from both pca and high energy x - ray timing experiment ( hexte ) were used for spectral fitting to obtain the fluxes for the all available xte j1807 - 294 _ rxte _ observations .
first , all the events collected by two event analyzers ( eas ) were combined ; then , the event arrival times were further corrected to the barycenter of the solar system , using jpl de200 ephemeris and the source position determined by _
observation @xcite .
only the events detected by the top layer of each pcu and with photon energies between 2 to 10 kev were selected in order to improve the sensitivity .
subsequently , all the selected events were divided into 150s data segments for further analysis .
the epoch folding search technique with a search resolution of @xmath15 sec was applied to test the significance of pulsation for each 150s data segment and to obtain the preliminary binary orbital parameters .
the test results show that the 190 hz pulsation can be detected only in the first @xmath090d ( up to tjd 12783 ) out of the overall @xmath0150d s observations . in order to correctly estimate the orbital parameters , only those data segments with the maximum @xmath16 value from epoch folding search larger than three times of the degrees of freedom (
dof=31 for 32 bins / period , false alarm probability @xmath17 ) were taken into account . the trial period with maximum @xmath16
was considered the best pulsar period for each data segment , and the corresponding error was evaluated using the equation proposed by @xcite .
the doppler effect owing to the orbital motion can be clearly observed with the best - fitted mean pulsar frequency @xmath18190.62353(3 ) hz ( or @xmath19 ms ) and binary orbital parameters : @xmath2040.0735(2 ) min , @xmath21 lt - ms and @xmath22tjd12720.68245(6 ) .
neither orbital eccentricity nor the spin frequency derivative can be significantly detected .
improved orbital parameters and the evolution of the pulsar spin frequency can be obtained from the pulse arrival time delay ; hence the above results were applied as the initial guess values for the following phase analysis .
irrespective of whether the complex spin change could be a result of the hot spot moving on the surface of the neutron star or accretion torque , we can expect the intrinsic phase ( after binary barycenter correction ) to be very complicated .
therefore , we employed an alternative method to obtain the precise orbital parameters instead of using a method analogous to the one proposed by @xcite which is applied to accretion millisecond pulsars ( e.g. @xcite ) . because the radial velocity amplitude of the neutron star ( @xmath23 ) in xte j1807 - 294 is small ( @xmath24 ) , the pulsation can be still detected in a 150-sec data segment if the folding frequency is close to the true spin frequency of the neutron star .
the pulse arrival time delay due to the orbital motion can be obtained from the pulse phase shift . the pulse profile folded with the mean pulsar period derived from the orbital doppler effect
is better fitted with a two - fourier component model ( fundamental + first overtone ) .
typical pulse profiles are shown in figure [ pulse_profile ] .
the peak of the fitted profile was chosen as the fiducial point for the following analysis .
the error of the phase for each 150s data segment was estimated from 1000 runs of monte carlo simulations .
in addition to the time delay due to the orbital motion , significant linear phase drift ( more than 1 cycle per day ) was detected in @xmath03.5 hours exposure because of slight inconsistency between the folding period and the neutron star spin period . to minimize the possible cycle count ambiguity , we adjusted the folding period to @xmath25 ms ( or @xmath26190.6235070 hz ) , based on the best pulsar ephemeris derived by the observation on tjd12700 , and applied it to the entire data . to test
if the intrinsic phase evolution is indeed complex , we fitted the observed pulse phase variation with the combined function of the time delay due to binary orbit plus evolution of intrinsic pulse phase modeled as a polynomial , that is @xmath27 however , no simple polynomial ( up to 8th order ) can yield acceptable fit ( reduced @xmath28 4.2 ) .
we conclude that the evolution of intrinsic pulse phase is too complicated to be modeled as a simple polynomial .
since further study of the nature of accreting millisecond pulsar requires precise binary orbital ephemeris , we alternatively tried to remove the intrinsic phase trend for daily observation .
although the evolution of intrinsic pulse phase was complicated for the first @xmath090d s of observations , the higher order terms of the intrinsic phase drift can be neglected for the daily observation because of the short exposure ( @xmath293.5 hrs ) .
therefore , we fitted a linear trend plus a single sinusoidal function due to the orbital motion with a fixed trial orbital period derived from the doppler effect and then removed only the linear trend from the phases .
the trend - removed phases were fitted with a circular orbit model to obtain updated orbital parameters . to further improve the orbital parameters ,
the new orbital period was used as the trial orbital period and above processes were repeated .
the iterations of this procedure continued until improvements of all parameters were at least an order smaller than the corresponding errors .
the final trend - removed phases are well fitted with a circular orbit whose best parameters are listed in table [ tb_op ] . the f - test shows that adding the eccentricity to the orbital model dose not significantly improve the fit .
the f value is 1.3 , which indicates that probability for improvement of the fit occurring by chance can be as high as @xmath030% .
the pulse arrival time delay due to the binary orbital motion is shown in figure [ orb_delay ] .
the pulse phase or spin frequency variation of the neutron star can be studies on basis of precise orbital parameters .
for obtaining the binary barycenter - corrected pulse phases ( hereafter , pulse phases ) , we folded the event time @xmath30 with a time - variable frequency due to the orbital doppler effect using the parameters ( @xmath31 , @xmath32 , and @xmath33 ) derived in section [ bop ] as @xmath34 dt \right\ } \\ \nonumber & = & frac \left\{\nu_0(t_i - t_0)-\nu_0{{a_x \sin i } \over { c}}\cos \left[{{2 \pi ( t_i - t_{\pi / 2 } ) } \over { p_{orb}}}\right]+\nu_0{{a_x \sin i } \over { c}}\cos \left[{{2 \pi ( t_0-t_{\pi / 2 } ) } \over { p_{orb}}}\right]\right\ } \end{aligned}\ ] ] where @xmath35 is the fraction of cycle counts calculated inside the braces , @xmath36 is folded frequency ( see section [ bop ] ) and @xmath37 is the phase zero epoch of the pulsation .
this operation is equivalent to first correcting the event times to the binary system barycenter and then folding with a constant frequency @xmath36 .
the pulse profile for each observation i d was constructed by through binning the event phases .
figure [ time_vs_phase_flux ] shows the pulse phase variation during the 2003 outburst . the pulse phases change like a smooth curve with several anomalous negative phase shifts occurring around march 9 ( tjd 12707 ) , 15 ( tjd 12713 ) , 26 ( tjd 12724 ) , 30 ( tjd 12728 ) and also around may 2(tjd 12761 ) with time scale of several days but become somewhat irregular after may 11 ( tjd 12770 ) ( see top panel of figure [ time_vs_phase_flux ] ) .
the first four anomalous phase shifts are coincident with the `` puny '' flares discovered by @xcite ( see figure 1 of their paper ) .
@xcite also reported that the pulsed fraction increases with the flares .
therefore , we examined the fractional root mean square ( rms ) amplitude for the first @xmath090d s observation .
figure [ pulse_fraction ] shows two additional significant enhancements in pulsed fraction around tjd 12761 and 12769 .
the former one is coincident with the fifth anomalous negative phase shift . to further investigate the coincidence of anomalous phase shifts and flares
, we made a 2 - 10 kev light curve for @xmath0150d observations through spectral fitting .
we used the archived pca standard-2 and hexte data and adopted the spectral model of absorbed blackbody plus a power law to obtain the 2 - 10 kev fluxes .
this model gives reasonable fit results for all observations with reduced @xmath38 of 0.6 - 1.6 .
the bolometric correction estimated by the mean ratio of the integrated 2 - 200 kev to 2 - 10 kev is @xmath39 .
figure [ time_vs_phase_flux ] shows the 2 - 10 kev flux of xte j1807 - 294 during 2003 outburst .
four strong flares with peak fluxes detected on tjd 12707 , 12713 , 12724 and 12728 , as well as two weak flares with peak fluxes detected on tjd12761 and 12769 , can be clearly seen on the light curve .
we labeled these flares from 1 to 6 according to their chronological order .
pulse phases exhibit negative phase shifts , and the pulsed fractions increase during the flare states . from figure [ time_vs_phase_flux ] and figure [ pulse_fraction ] , the flares are likely to appear in pairs with peak separations of 5.9d for flares 1 and 2 , 4.8d for flares 3 and 4 and 7.8d for flares 5 and 6 .
these negative phase shifts are more likely due to the motion of the accretion footprint rather than the spin - down torque acting on the neutron star , which will be discussed in section [ dis_ps ] .
further analysis results for the pulse behaviors during the flare states are presented in section [ edpb ] . on the other hand ,
the smooth pulse phase evolution for the non - flare state during the first @xmath060d of the outburst is possible owing to the accretion torque .
we defined the epochs for the flare states : tjd 12703 to 12716 for flares 1 and 2 , tjd 12720 to 12730 for flares 3 and 4 and tjd 12759 to 12778 for flares 5 and 6 . to investigate the smooth pulse phase evolution during the non - flare states
, we removed the phases during the flare states and also discarded the ones after tjd 12778 .
the non - flare phases were fitted with a polynomial , starting from a quadratic function and then subsequently adding higher order terms .
the f - test was applied to verified if the additional higher order term significantly improves the fitting ( with confidence level @xmath40 90% ) .
we found that the phase evolution for the non - flare states is better described by a 4th order polynomial ( see figure [ phase_ev ] ) . on the other hand , imperfect solar system barycenter correction due to
source position error results in correction error for event arrival time @xcite , which further leads to systematic errors in spin parameters . from eq . 5.3 in @xcite ,
the systematic errors of the pulse phase and spin parameters , including phase zero epoch ( @xmath37 ) , spin frequency ( @xmath41 ) and its derivatives ( @xmath42 ) can be expressed as @xmath43 \\ \delta t_0 & = & { { \delta \phi } \over \nu } \approx { { r_e \delta \theta}\over { c } } \sin[\omega_e(t - t_\gamma)-\lambda-\beta ] \\ \delta \nu & = & { { d ( \delta \phi ) } \over dt}\approx \omega { { r_e \nu \delta \theta}\over { c } } \cos[\omega_e(t - t_\gamma)-\lambda-\beta ] \\ \delta
\bigl({{d^n \nu } \over { dt^n}}\bigr ) & = & { { d^{n+1 } ( \delta \phi ) } \over dt^{n+1}}\end{aligned}\ ] ] where @xmath44 is the orbital radius of the earth , @xmath45 is earth s orbital angular velocity , and @xmath46 is the time of vernal point . the best fitted spin parameters and the corresponding statistic and systematic errors are listed in table [ tb_sp ] .
the frequency derivative @xmath47 derived from the best fitted spin parameters shows that the neutron star was spun - up during the first @xmath060d of the 2003 outburst ( see figure [ phase_ev ] ) .
further , the non - flare light curve can be well described as the combination of a fast and a slow exponential decay component with decay time constants of @xmath48 d and @xmath49 d respectively ( see figure [ time_vs_phase_flux ] ) .
@xcite discovered that a high - energy pulse leads a low - energy pulse by about 200 @xmath14s ( referred to as `` soft lags '' ) from 2 to 10 kev but saturates at a higher energy for the accreting millisecond pulsar sax j1808.8 - 3658 .
this energy - dependent pulse arrival time has been well explained by the scenario proposed by @xcite and @xcite .
the emissions mainly consist of a soft blackbody component from a heated spot on the neutron star surface and a hard component from the comptonization of the blackbody photons ( seed photons ) in the plasma heated by an accretion shock slab with thomson optical depth of 0.3 to 1 .
the hard comptonization component leads the soft blackbody component as a result of the different angular distribution of the soft and hard components and the doppler boosting effect .
the low - energy pulses lag behind the high - energy pulses because the latter contain relative less soft blackbody component . for the photons
energies greater than @xmath010 kev , only the comptonization component is left so the hard leads saturate .
the emission spectra of many accretion - powered millisecond pulsars are well described by this two - component model , including xte j1807 - 294 @xcite .
subsequent to the discovery of soft lags in sax j1808.8 - 3658 , similar phenomenon was also seen in some of accreting millisecond pulsars , such as xte j1715 - 305 @xcite and hete j1900.1 - 2455 @xcite .
however , for igr j00291 + 5934 , the soft lags increase up to 7 kev and then decrease at higher energy bands instead of saturation @xcite .
the decrease at higher energy bands is likely due to the time lags caused by compton upscattering @xcite . to investigate the energy - dependent pulse behaviors for xte j1807 - 294
, we divided 2 - 20 kev photons into 8 energy bands for the first @xmath060d observations .
since we assumed hot spot to be fixed during the non - flare state , we folded the non - flare event times of each energy band with a ephemeris ( hereafter , the best ephemeris ) from the variable observed pulse frequencies due to the orbital doppler effect and neutron star spin parameters from the 4th order polynomial fitting ( see table [ tb_sp ] ) .
thus , pulse phases of the non - flare state are expected to be a constant with time .
the average pulse profile for the non - flare state of each energy band was obtained by binning the corresponding event phases .
the energy - dependent pulse arrival time delays , measured relative to the pulsation of the softest band ( 2 - 3.6 kev ) , were obtained through the cross correlation of the pulse profiles .
the high energy pulses lead soft pulses up to @xmath50 @xmath4 ( @xmath51 cycle ) from 2 to 10 kev and the leads saturate for 10 to 20 kev pulse ( figure [ soft_lag ] ) , similar to that seen in sax j1808.8 - 3658 @xcite . on the other hand ,
the fractional rms pulse amplitude slightly increases from 4.5% to 5% in the 2 - 4 kev range , remains constant to 7.5 kev , and largely decreases to only 2% at 20 kev ( see figure [ pulse_frac_band ] ) .
significant phase shift , and its correlation to the flux and fractional pulse amplitude , has never observed for other accreting millisecond pulsars . to further understand its origin , we studied the energy - dependent phase shifts during the flare state .
the pulse phase defined here is identical to the one in section [ sl ] , that is , folded with the best ephemeris .
the phase shift for each energy band , defined as the phase difference between the flare state and the corresponding mean phase of the non - flare state , was obtained through the cross correlation method . however , since this ephemeris is applicable for the first @xmath060d data , only the phase shifts of the first four flares can be well - defined .
figure [ phase_shift_band ] shows that the phase shifts of different energy bands when the fluxes of flares attained their maximum values for the flares 1 to 4 .
significant phase shifts can be observed in all energy bands for all the four flares . in general , the phase shifts decrease for higher energy bands from 2 to 5 kev , almost a constant between 5 - 11 kev and slightly increase for the hardest energy band .
we have derived the precise binary orbital parameters for the x - ray binary xte j1807 - 294 .
the orbital parameters have been previously reported by @xcite , @xcite , @xcite , and @xcite , whose values are listed in table [ tb_opc ] for comparison .
all the parameters are consistent with each other except the epoch of the mean longitudes for the binary orbit ( i.e. phase zero epochs ) .
the epoch of mean longitude is chosen as when mean longitude=@xmath52 ( @xmath53 ) in this paper . @xcite claimed that they used same reference point , but it deviates from ours by about @xmath00.5 phase .
we examined their arguments and found that they probably used the positive - crossing time in time vs. pulsar period plot ( see figure 6 in @xcite ) , which is equivalent to the mean longitude=@xmath54 , as the phase zero epoch . if their reference time is indeed @xmath55 instead of @xmath53 , the value proposed by them and the one in this paper are consistent . on the other hand ,
the epoch of mean longitude considered by @xcite is equivalent to the mean longitude=@xmath56 , which is close but still significantly different from what we found ( more than 8@xmath57 ) . their value , and
the one proposed by @xcite , is probably due to a short observation time span and , as mentioned in @xcite , the complex method used by these authors . with the precise orbital parameters
, we have found significant pulse phase shifts on the time scale of one to several days coincident with the flares as well as pulsed fraction enhancements . if the hot spot is assumed to be fixed on the neutron star surface and the pulse phase variation
truly reflects the history of the change in the spin frequency of the neutron star , fitting a quadratic curve for the pulse phases around the flare states implied a spin - down rates @xmath58 hz / s .
it is much larger than the theoretical expectation value ( @xmath59 hz / s , see @xcite ) or the reported values for accretion - powered millisecond pulsars ( @xmath60 hz / s , see e.g. @xcite ) .
in addition , if the flares are caused by an enhanced accretion , the additional accretion torque should further spin - up the neutron star rather than spin it down @xcite .
thus , the flare - coincident negative phase shifts are highly unlikely due to the accretion torque .
another possible explanation for the anomalous pulse phase shifts is the hot spot moving on the surface of neutron star .
assume that the neutron star is rotating eastward as the earth and that the hot spot is fixed at a position in the northern hemisphere during the non - flare state .
as the accretion rate increases in a flare state , the hot spot moves eastward , in the same direction as that of the neutron star rotation , so that the pulse peak arrives to an observer earlier than the one expected for the non - flare state and results negative phase shift . at the time when the flare reaches its maximum
, this movement can be as large as about 70@xmath61 ( @xmath00.2 cycle ) in longitude .
when the flare s flux decreases , the hot spot moves westward and eventually returns to its original location at the end of the flare .
the correlation between phase shift and flux has been suggested for other accreting millisecond pulsars , although the no flare similar to that observed in xte j1807 - 294 has ever been detected in other accreting millisecond pulsars .
@xcite reported the discovery of high anti - correlation between x - ray flux and the pulse phase jitter ( @xmath62 ) around the mean quadratic phase trend in the accreting millisecond pulsar xte j1814 - 338 .
such a correlation is very similar to what we found in xte j1807 - 294 during the flare states although the correlation between the pulsed fraction and the flux is still unknown for xte j1814 - 338 .
recently , @xcite discovered an anomalous large phase jump for sax 1808.4 - 3658 during its 2002 outburst , which is also likely as a result of a large accretion rate change ( see figure 1 in @xcite ) .
@xcite suggested that the flares seen in xte j1807 - 294 could result from accreting blobs of high mass density , lasting for hours to days , at the inner edge of the accretion disk .
the positive correlation of khz qpo frequency and pulsed fraction implies that the surface emission properties of neutron star are very sensitive to the inhomogeneous disk flow @xcite .
we found that not only the pulsed emissions increase but also the position of the hot spot moving on neutron star surface .
since the longitudinal movement of the hot spot on the neutron star results in the phase shift , it is also possible that the hot spot moves in latitude , which could be a part of the reason for the pulse amplitude enhancements during the flare states .
the flux and phase shift coincidence has also been seen in accreting pulsars with much lower spin frequency .
@xcite reported that during the bursting x - ray pulsar gro j1744 - 28 ( @xmath63hz ) type ii bursts , likely a consequence of accretion instability , the pulse arrival time lags by @xmath64 ms relative to the non - burst state .
pulsed fraction is also strongly enhanced during the bursts @xcite .
@xcite proposed a scenario to describe the pulse phase shifts .
because of the misalignment of the neutron star spin axis and magnetic dipole moment , the accretion flow is asymmetric and its footprint forms an arc surrounding the magnetic pole .
the enhanced accretion causes the footprint to move around magnetic pole as a consequence of the deformation of the field lines .
the azimuth movement of the footprint with respect to the spin axis results in the observed phase shift . except that the phase shift is negative for xte j1807 - 294 but positive ( pulse arrival delay ) for gro j1744 - 28 and the different time scales for flare ( several days ) and type ii burst ( several seconds ) ,
the coincidence of phase shift , pulse amplitude enhancement and burst in gro j1744 - 28 is very similar to what we report for xte j1807 - 294 .
the scenario suggested by @xcite can probably give us a clue to explain the observed behaviors of xte j1807 - 294 .
form the discovery of the coincidence between the flare and phase shift in xte j1807 - 294 , as well as the correlation between the x - ray flux and the pulse phase jitters in xte j1814 - 338 @xcite , hot spot position on accreting millisecond pulsar is sensitive to accretion flow , as suggested by @xcite .
although we assume that the accretion hot spot is still fixed on the surface neutron star during the non - flare state , we can not exclude the possibility that the hot spot marginally drifts as the x - ray flux changes .
therefore , it is difficult to determine whether the pulse phase variation is a consequence of the spin frequency change or hot spot drift on the neutron star surface .
if the correlation between the hot spot drift and accretion flow is applicable to all accreting millisecond pulsars , the measurements of the spin parameters ( e.g. @xmath65 ) from which accretion torque is estimated , are not reliable , including the ones listed in table [ tb_sp ] for xte j1807 - 294 .
this may also explain the large disparity between the measured spin frequency derivative and the theoretical value calculated form the accretion torque reported by @xcite for sax j1808.4 - 3658 in its 2002 outburst .
[ [ section ] ] 1 . + + due to complicated intrinsic phase variations , to obtain the precise orbital parameter , we folded our data sets with a constant frequency and used the trend removal technique to remove the local linear pulse phase trend from the observed pulse phase variation for daily observations .
we obtained more precise orbital parameters consistent with the ones previously reported ( except for @xmath66 ) .
the binary barycenter - corrected pulse phases change smoothly except for several anomalous negative phase shifts coincident with flux flares and pulse amplitude enhancements .
the pulse fractional amplitude variation indicates that there were six flares during the first @xmath090d of the xte j1807 - 294 in its 2003 outburst .
the phase changes of non - flare pulse imply that the neutron star was spun - up during the first 60d of the outburst under the assumption that the hot spot was fixed on the surface of the neutron star during the non - flare states .
the analysis results of the energy - dependent pulse arrival time indicates that hard pulses lead soft pulses up to @xmath50 @xmath4 from 2 to 10 kev , and the lead saturates from 10 to 20 kev , which agrees with the scenario proposed by @xcite and @xcite .
our discovery of the coincidence of flare and phase shift , as well as its correlation with pulse amplitude and khz qpo @xcite in xte j1807 - 294 , implies that probably due to weak magnetic confinement on the accreting material , the hot spot position on the accreting millisecond pulsars is sensitive to the accretion flow .
the authors thank r. e. taam , j. h. swank , h. c. spruit , m. c. miller and d. lai for helpful discussions .
this research has made use of data obtained through the high energy astrophysics science archive research center ( heasarc ) , provided by the nasa s goddard space flight center .
this work is supported by the grants from national science council nsc 94 - 2112-m-008 - 003 and nsc 95 - 2112-m-008 - 026-my2 . )
( top panel ) and 2 - 10 kev flux ( bottom panel ) .
a total six flares were detected during the flux declination .
the arrows in both panels indicate the flare peak time .
clear flare and phase shift coincidence can be seen .
the solid line on the bottom plot represents the best fitted two - component exponential model for the non - flare state fluxes.[time_vs_phase_flux ] ] cc + parameter & value + orbital period , @xmath31 ( min ) & 40.073601(8 ) + projected semimajor axis , @xmath67 ( lt - ms ) & 4.823(5 ) + epoch of 90@xmath61 mean longitude , t@xmath68 ( tjd / tdb ) & 12,720.682574(6 ) + mass function ( @xmath69 ) & @xmath70 + eccentricity , @xmath71 & @xmath720.006(2@xmath57 ) + @xmath38/degree of freedom & 2171/1880 + cccc + parameter & value & statistic & systematic + & & error & error + phase zero epoch , @xmath37 ( tjd / tdb ) & 12700.00000000(2 ) & @xmath73 & @xmath74 + frequency at @xmath37 , @xmath36 ( hz ) & 190.62350686(8 ) & @xmath75 & @xmath76 + @xmath77 at @xmath37 ( hz / s ) & @xmath78 & @xmath79 & @xmath80 + @xmath81 at @xmath37 ( hz / s@xmath82 ) & @xmath83 & @xmath84&@xmath85 + @xmath86 ( hz / s@xmath87 ) & @xmath88 & @xmath89&@xmath90 + @xmath38/degree of freedom & 22.76/26 + lllllc + mission & @xmath31 & _ a _ sin_i _ & epoch of mean & phase for epoch & reference + & ( min ) & ( lt - ms ) & longitude(tjd ) & of mean longitude & + & 40.0741(5 ) & - & - & - & 1 + _ xmm - newton_&40.0741(fixed)&@xmath91&12720.68644(13 ) & 0.140(5 ) & 2 + _ xmm - netwon_&40.0741(fixed)&@xmath92&12720.67415(16)&0.699(6 ) & 3 + _ xmm - netwon_&40.0741(fixed)&@xmath93&12720.668(2)&0.48(7 ) & 4 + _
rxte_&40.073601(8)&@xmath94&12720.682574(6 ) & - & 5 + | this study reports pulse variation analysis results for the forth discovered accretion - powered millisecond pulsar xte j1807 - 294 during its 2003 outburst observed by _ rossi x - ray timing explorer_. the pulsation is significantly detected only in the first @xmath090d out of @xmath0150d observations .
the pulse phase variation is too complex to be described as an orbital motion plus a simple polynomial model .
the precise orbital parameters with @xmath1 min and @xmath2 lt - ms were obtained after applying the trend removal to the daily observed 150s segments pulse phases folded with a constant spin frequency without keplerian orbit included .
the binary barycenter corrected pulse phases show smooth evolution and clear negative phase shifts coincident with the flares seen on the light curve and the enhancements of fractional pulse amplitude .
the non - flare pulse phases for the first @xmath060d data are well described as a fourth order polynomial implying that the neutron star was spun - up during the first @xmath060d with a rate @xmath3 hz / s at the beginning of the outburst .
significant soft phase lags up to @xmath0500 @xmath4 ( @xmath010% cycle ) between 2 to 20 kev were detected for the nonflare pulse phases .
we conclude that the anomalous phase shifts are unlikely due to the accretion torque but could result from the `` hot spot '' moving on the surface of neutron star . |
the minimal @xmath0 extension of the standard model ( sm ) , which is based on the gauge group @xmath2 , provides a satisfactory explanation for the non - zero neutrino masses @xcite . in this class of models ,
sm singlet fermions are naturally introduced in order to cancel the associated anomaly .
these particles are accounted for right - handed neutrinos and hence a seesaw mechanism can be obtained .
it was shown that light neutrino masses can be generated within @xmath0 extension of the sm through either type - i seesaw @xcite or inverse seesaw mechanism @xcite . in type - i seesaw mechanism
right - handed neutrinos acquire majorana masses at the @xmath0 symmetry breaking scale , therefore the neutrino s yukawa coupling must be @xmath3 , while in inverse seesaw these majorana masses are not allowed by the @xmath0 gauge symmetry and another pair of sm gauge singlet fermions with tiny masses @xmath4 kev must be introduced .
one of these two singlets fermions couples to right handed neutrino and is involved in generating the light neutrino masses .
furthermore , it was shown that in a susy context , the @xmath0 and susy scales can be correlated through the mechanism of radiative breaking of @xmath0 symmetry , similarly to the radiative electroweak symmetry breaking in mssm @xcite . in particular , it was proven that the radiative corrections in @xmath0 extension of the mssm ( blssm ) with type i seesaw may drive the squared mass of extra higgs boson from positive initial values at the gut scale to negative values at the tev scale , leading to spontaneous breaking of @xmath0 .
thus , the energy scale of @xmath0 breaking is naturally related to the susy breaking scale .
however , it was pointed out @xcite that the breaking of @xmath0 in this model depends crucially on the large value of the right - handed neutrino yukawa coupling and it is possible to break the @xmath0 through the vacuum expectation value ( vev ) of the right - handed sneutrino . in this case
@xmath1-parity is also spontaneously broken and the resulting model will have quite involved phenomenology . in this paper we analyze the radiative @xmath0 symmetry breaking in blssm with inverse seesaw ( blssm - is ) .
we show that the breaking of @xmath0 occurs for a wider region of parameter space through the vev of the higgs singlet .
we consider the renormalisation group equations ( rges ) to show explicitly that for wide range of parameters the squared mass of the higgs singlet can be negative at tev scale while the squared mass of the right - handed sneutrino remains positive .
therefore , the @xmath0 symmetry is spontaneously broken by the vev of this singlet and @xmath1-parity remains exact .
in addition , using the program of vevacious @xcite , we analyze the vacuum stability in both blssm - is and blssm - type i. we show that , unlike the blssm - type i , in blssm - is the vev of right - handed sneutrino is always close to zero and much less than the vev of the singlet scalar that breaks the @xmath0 and keeps @xmath1-party conserved .
the plan of the paper is as follows . in the next section ,
we analyze the rge running and the radiative @xmath0 symmetry breaking in blssm with inverse seesaw and compare it with the results of the blssm with type i seesaw . in section 3
we investigate the vacuum stability in the blssm - is and also in blssm - type i. we conclude in section 4 .
tev scale blssm - is is based on the gauge group @xmath5 , where the @xmath6 is spontaneously broken by chiral singlet superfields @xmath7 with @xmath0 charge @xmath8 as in conventional @xmath0 model , a gauge boson @xmath9 and three chiral singlet sueperfields @xmath10 with @xmath0 charge @xmath11 are introduced for the consistency of the model .
finally , three chiral singlet superfields @xmath12 with @xmath0 charge @xmath13 and three chiral singlet superfields @xmath14 with @xmath0 charge @xmath15 are considered to implement the inverse seesaw mechanism @xcite .
the superpotential of the leptonic sector of this model is given byw=
y_e e^c l h_1 + y _ _ r^c lh_2 + y_s _ r^c _ 1 s_2 + h_1h_2 + _ 1_2 . [ sp ] note that the chiral singlet superfields @xmath16 and @xmath17 have the same @xmath0 charge .
therefore , one may impose a discrete symmetry in order to distinguish them and to prohibit other terms beyond those given in eq .
( [ sp ] ) . in this case , the relevant soft susy breaking terms , assuming the usual universality assumptions , are as follows - l_soft & = & _ _ ^2 ^2 + y_^a_r^c h_2 + y_e^a^ch_1 + y_s^a_r^cs_2 _ 1 + b h_1 h_2 + b _ 1 _ 2 + & + & m_1 + m_2^a ^a + m_3 ^a ^a + m_bl _
bl_bl+ h.c , [ soft ] where the sum in the first term runs over @xmath18 and @xmath19 ( @xmath20 ) is the trilinear scalar interaction coupling associated with lepton yukawa coupling . in order to prohibit a possible large mass term @xmath21 in the above
, we assume that the particles , @xmath22 , @xmath7 , and @xmath14 are even under matter parity , while @xmath12 is an odd particle . the @xmath0 symmetry can be radiatively broken by the non - vanishing vacuume expectation values ( vevs ) @xmath23 and @xmath24 @xcite . the tree level potential @xmath25 is given by @xmath26 where @xmath27 and @xmath28 .
the stablitity condition of @xmath29 is given by @xmath30 a non - zero minimum may be obtained if there is a negative squared mass eigenvalue in the @xmath0 higgs mass matrix , _
i.e. _ , if @xmath31 this condition is not satisfied at the gut scale with universal soft breaking terms .
however , as we will show , similar to the mssm scalar higgs masses , the running from a large scale down to tev scale , @xmath32 and @xmath33 will have different renormalization scales so that the minimization condition is eventually satisfied , and hence , the @xmath0 symmetry is spontaneously broken .
the minimization conditions , @xmath34 , lead to the following equations : @xmath35 where @xmath36 and @xmath37 .
these two equations are similar to the electroweak symmetry breaking conditions in mssm which are used to determine the value of @xmath38 and @xmath39 parameters at the electroweak scale .
it is worth noting that in mssm , where @xmath40 , one can not satisfy the condition of non - vanishing vevs and @xmath41 unless the running from gut to weak scale reduces @xmath42 to negative values , thanks to the large yukawa coupling of @xmath43 with top quark .
the situation with @xmath0 symmetry breaking could be different .
the conditions @xmath44 and @xmath45 can be simultaneously satisfied with positive @xmath46 , if @xmath47 . before elaborating this point , let us consider the running of the scalar masses @xmath46 and also the right - handed sneutrino squared masses , @xmath48 , via the @xmath0 renormalization group equations ( rges ) in both type i and inverse seesaw mechanisms .
in type i seesaw , these rges are given by 16 ^ 2 & = & -12 g_bl^2 m_bl^2 + 4 m__1 ^ 2 tr(y__r y__r^ * ) + 4 tr(y^a^*__r y^a__r ) + 8 tr(m^2__r y__r y^*__r ) , + 16 ^ 2 & = & -3 g_bl^2 m_bl^2 + 8 m__1 ^ 2 y__r y__r^ * + 8 y^a^*__r y^a__r + 4 m^2__r y__r y__r^ * + 8 y__r m^2__r y__r^ * + 4 y__r y__r^ * m^2__r , where @xmath49 is the yukawa coupling of right - handed neutrino term in type
i superpotential : @xmath50 and the trilinear coupling @xmath51 is defined as usual as @xmath52 .
thus for @xmath53 one finds 16 ^ 2 & = & -12 g_bl^2 m_bl^2 + 4 y^2__r ( m__1 ^ 2 + a__r^2 + 2 m^2__r ) , + 16 ^ 2 & = & -3 g_bl^2 m_bl^2 + 8 y^2__r ( m__1 ^ 2 + a__r^2 + 2 m^2__r ) .
the last term proportional to @xmath49 in these equations derives the mass squared negative at tev scale .
therefore , is clear that @xmath48 can be negative before @xmath54 ( due to the large coefficient of @xmath49 in the rge of @xmath48 ) . in this case of hierarchal @xmath49 , both @xmath0 and @xmath1-parity
will be spontaneously broken @xcite . however , in case of @xmath55 , the equations take the form 16 ^ 2 & = & -12 g_bl^2 m_bl^2 + 12 y^2__r ( m__1 ^ 2 + a__r^2 + 2 m^2__r ) , + 16 ^ 2 & = & -3 g_bl^2 m_bl^2 + 8 y^2__r ( m__1 ^ 2 + a__r^2 + 2 m^2__r ) .
therefore it is expected that @xmath54 becomes negative at tev scale while @xmath48 remains positive , so @xmath0 symmetry is spontaneously broken and @xmath1-parity remains exact @xcite . in inverse seesaw ,
the relevant rges are given by 16 ^ 2 & = & -12 g_bl^2 m_bl^2 + 2 m__1 ^ 2 tr(y_s y_s^ ) + 2 tr(y^a^*_s y^a^t_s ) + 2 tr(m^2__2 y^_s y_s ) + 2 tr(m^2__r y_y^ _ ) , + 16 ^ 2 & = & -12 g_bl^2 m_bl^2 , + 16 ^ 2 & = & -3 g_bl^2 m_bl^2 + 2 m__1 ^ 2 y_s y_s^+ 2 y^a_s y^a^_s + 4 y^a _
y^a^_+ m^2__r y_s y_s^+ 2 m^2__r y_y_^ + & + & 2 y_s m^2__2 y_s^+ y_s y_s^m^2__r + 2 y_y_^m^2__r , + 16 ^ 2 & = & -3 g_bl^2 m_bl^2 + 2 m__1 ^ 2 y_s^y_s + m^2__2 y_s^y_s + 2 y^a^_s y^a_s + 2 y^_s m__r^2 y_s+ y_s^y_s m^2__2 .
thus for hierarchical yukawas : @xmath56 and @xmath57 , one gets 16 ^ 2 & = & -12 g_bl^2 m_bl^2 + 2y_s^2 ( m__1
^ 2 + a_s^2 + m__2 ^ 2 ) + 2 m__r^2 y_^2 , + 16 ^ 2 & = & -3 g_bl^2 m_bl^2 + 2 y_s^2 ( m__1 ^ 2 + a_s^2 + m^2__2 + m__r^2 ) + 4 y_^2 ( m__r^2 + a_^2 ) , + 16 ^ 2 & = & -3 g_bl^2 m_bl^2 + 2 y_s^2 ( m__1 ^ 2 + m__2 ^ 2 + a_s^2 + m__r^2 ) . from these equations
, one can see that in inverse seesaw scenario the evolution of @xmath54 and @xmath48 depends on the relative strength of @xmath58 and @xmath59 . in the case of hierarchal yukawa ,
@xmath54 can be of order or slightly smaller than @xmath48 if @xmath60 . from the rges of @xmath58 and @xmath59
one can notice that @xmath59 must be @xmath61 , to avoid a possible landau pole at high scale , while @xmath58 can be of order one . in these conditions ,
the masses @xmath54 and @xmath48 are of the same order and positive ( as shown explicitly in fig .
[ rge - blssmis ] ) .
so that @xmath1-parity remains exact symmetry and @xmath0 could be broken if symmetry breaking conditions in eq .
[ tadapole2 ] are satisfied . as intimated , these conditions do not require negative mass squared scalar masses and with @xmath62 , symmetry can be broken with @xmath63 .
so even if @xmath64 and both are positive , @xmath0 symmetry only can be broken . in case of degenerate yukawa , _
i.e. _ , @xmath65 and @xmath66 , the splitting between @xmath54 and @xmath67 becomes larger , so that @xmath54 can be negative while @xmath48 and @xmath68 are positive . + in fig .
[ rge - blssmis ] we display the scale evolution of the higgs masses @xmath46 and also the scalar masses @xmath69 and @xmath70 based on the numerical solution of complete rges derived by using sarah @xcite , for @xmath71 tev and @xmath72 is assumed . as can be seen from this figure
, @xmath73 drops rapidly to the negative region , while @xmath74 and other scalar masses remain positive at tev scale .
analogously to the radiative electroweak symmetry breaking in mssm , this mechanism works with large yukawa coupling .
it is worth mentioning that unlike the type i seesaw blssm , here the scalar mass @xmath69 remains positive at the low scale independently of the initial values .
hence the @xmath0 breaking via a non - vanishing vev for right - handed sneutrinos @xmath75 , does not occur in the present framework and @xmath1-parity remains exact . as intimated ,
if right - handed sneutrino acquires a non - vanishing vev , then both @xmath0 and @xmath1-parity would simultaneously be broken . in this case
, the model leads to a quite different and involved phenomenology at the low scale @xcite .
in this section we analyze , by using vevacious @xcite , the vacuum stability of the blssm - is .
we perform a wide scan over all relevant parameters and calculate the scalar potential through sarah @xcite and spheno @xcite .
the stability results are classified into two categories : @xmath76 non - vanishing vevs of @xmath77 : @xmath78 gev and @xmath79 , where @xmath0 is spontaneously broken and @xmath1-parity is conserved .
@xmath80 @xmath81 and @xmath82 gev , so that both @xmath0 and @xmath1-parity are spontaneously broken . in ref .
@xcite , it was shown that in blssm with type i seesaw , out of more than 2000 scanned points only about 100 points may lead to global non - zero vevs for neutrinos that may break the @xmath1-parity , while all other points preserve @xmath1-parity and break @xmath0 only .
we confirmed these results by considering a wider range of parameter space .
our results are presented in fig .
[ blssm - vevs ] , where the vevs : @xmath83 and @xmath84 are given in terms of the relevant parameters @xmath85 , @xmath86 and the gauge coupling mixing , @xmath87 , between @xmath88 and @xmath6 .
we also display the correlation between @xmath83 and @xmath84 . as can be seen from these plot ,
although most of the considered points lead to a non - vanishing @xmath83 and zero @xmath84 , _ i.e. _ , @xmath0 is spontaneously broken while @xmath1-parity remains exact , there is a non - negligible number of points which induce non - vanishing @xmath84 , hence @xmath1-parity is broken along with the @xmath0 .
it is noticeable that the possibility of obtaining non - vanishing @xmath84 is increased with large values of @xmath89 as found in previous section by the rge evolution .
in addition , the last plot in fig.[blssm - vevs ] clearly shows that for most of the parameter space one gets @xmath90 tev and @xmath91 .
nevertheless , it is quite plausible to have @xmath92 and larger than @xmath83
. those points will be the benchmarks of @xmath1-parity violation scenario studied in ref .
@xcite .
+ -1 cm we performed a similar analysis for the blssm - is .
we scanned over a large region of parameter space and checked the vevs of the scalar fields @xmath93 and @xmath75 .
[ blssmis - vevs ] shows the resulting vevs as functions of @xmath87 , @xmath86 and @xmath85 .
also the correlation between @xmath83 and @xmath84 is given in the last plot in this figure .
it is now clear that unlike the case of blssm with type i seesaw , in blssm - is there is no chance of getting non - zero vev for the right - handed sneutrinos and one always finds @xmath94 with @xmath95 .
this conclusion is independent of the values of @xmath85 , @xmath87 or @xmath89 as shown in fig .
[ blssmis - vevs ] .
moreover , the correlation between the vevs in the last plot confirms that @xmath96 can be order tev while @xmath97 vanishes identically .
therefore , we can conclude that in the blssm - is the @xmath0 symmetry can be radiatively broken while the @xmath1-parity remains as an exact symmetry .
we have analysed the radiative symmetry breaking of @xmath0 within the framework of blssm - is .
we considered the rges to show that for a wide range of parameters the squared mass of the higgs singlet can be negative at tev scale while the squared mass of the right - handed sneutrino remains positive .
therefore , the @xmath0 symmetry is spontaneously broken by the vev of this singlet and @xmath1-parity remains exact .
we also investigated the vacuum stability of the blssm - is , using the program of vevacious .
we showed that for a wide region of parameter space the singlet scalar @xmath93 gets a non - vanishing vev @xmath98 tev and @xmath99 so that @xmath0 is spontaneously broken and @xmath1-parity is conserved .
this conclusion is different from the results obtained in blssm - type i , where @xmath1-parity can be spontaneously broken for a non - negligible number of points in parameter space .
i would like to thank a. hammad for his valuable help in using the vevacious code and c. un for fruitful discussions .
this work was partially supported by the stdf project 13858 and the grant h2020-msca - rise-2014 n. 645722 ( nonminimalhiggs ) . | we study the radiative symmetry breaking of @xmath0 in supersymmetric models with inverse seesaw mechanism .
we show that for a wide region of parameter space the radiative corrections can drive the squared mass of the extra higgs boson from positive initial values at the gut scale to negative values at the tev scale , leading to the spontaneous breaking of the @xmath0 symmetry .
we also emphasize that in this class of models , unlike the supersymmetric @xmath0 models with type i seesaw , the right - handed sneutrino can not get a non - zero vacuum expectation value .
therefore , @xmath0 can be radiatively broken while @xmath1-parity remains an exact symmetry . |
a combined system is said to be entangled when its state can not be expressed as a product of states corresponding to the individual sub systems @xcite .
the entangled systems have interesting properties such as non - locality and contextuality which make them a great resource for various quantum protocols @xcite .
one generally uses the entanglement between two spatially separated particles in the same degree of freedom such as spin or polarization .
however , one can also have hybrid entanglement in which two degrees of freedom of a single particle or two particles are entangled @xcite .
this arises due to the non - separability of two degrees of freedom .
however , it is not an exclusive property of a quantum system .
similar kind of non - separability can be seen in classical optics , for example radially polarized light beams @xcite .
this quantum like classical entanglement has been receiving a lot of attention in recent years @xcite .
these non - separable states of light are shown to violate bell like inequality @xcite . furthermore , they find applications in polarization metrology and ultra sensitive angular measurements @xcite .
recently , it has been shown that phase singular beams or optical vortices also violate bell s inequality for continuous variables such as position and momentum @xcite .
these optical vortices carry an orbital angular momentum ( oam ) of @xmath0 per photon , @xmath1 being the azimuthal index or order of the vortex @xcite .
this oam can be used as an additional degree of freedom along with the polarization to form a hybrid entangled state that violates the bell s inequality for discrete variables @xcite .
scattering of structured light beams such as optical vortices has been studied for their coherence properties and applications @xcite .
it has been shown that one can generate partially coherent ring shaped beams from the scattering of coherent optical vortices @xcite . here , we generate light beams with non - separable oam and polarization and verify the preservation of non - separability under scattering through a rotating ground glass ( rgg ) .
these non - separable beams can be generated using q - plates @xcite or interferometers @xcite . in our set up
, we modify a polarizing sagnac interferometer @xcite to generate the non - separable beams by replacing dove prism with a spiral phase plate ( spp ) .
the generated beams scatter through a rgg and the scattered light is collected by a plano - convex lens to measure their polarization and intensity distributions at the focus .
we measure the degree of polarization of the beam , as a measure of non - separability @xcite , before and after scattering which should be @xmath2 for a maximally non - separable state and @xmath3 for a completely separable state .
we also project the scattered as well as coherent light to different polarizations and record the corresponding intensity distributions which confirm the non - separability . using the same experimental setup , we vary the degree of non - separability by controlling the intensities in the two arms of the interferometer . in section [ sec.2 ] we give a theoretical background to the oam - polarization non - separable state and describe the methods we used to witness the non - separability . experimental setup to generate the described states
is given in section [ sec.3 ] .
the results and discussion are given in section [ sec.4 ] and finally we conclude in section [ sec.5 ] . for simplicity , we use the dirac notation to describe the states even though we are using classical light beams .
a maximally entangled / non - separable state of polarization and oam can be written as @xmath4 where @xmath5 and @xmath6 are basis vectors of 2d complex vector spaces corresponding to the polarization and the oam subspace respectively .
we work in the paraxial domain with linear optics , where polarization and oam are independent .
thus \{@xmath5 } and \{@xmath6 } form two mutually independent complex vector spaces .
the density matrix for the non - separable state @xmath7 is given by @xmath8 .
one can obtain the reduced density matrix corresponding to the polarization @xmath9 by taking a partial trace of this density matrix over oam states , @xmath10 here , @xmath11 is a @xmath12 identity matrix . for a given density matrix @xmath13 describing a state in @xmath14 dimensional hilbert space
, one can define linear entropy @xcite @xmath15 @xmath16 characterizes the amount of mixedness for a given density matrix .
it is known that for an entangled / non - separable state , the subsystems will be in a mixed state .
stronger the non - separability , larger the amount of mixedness present in the subsystems .
thus by measuring linear entropy @xmath16 of the subsystem , one can measure the degree of entanglement or the non - separability .
for the maximally non - separable state given in eq .
[ 1 ] , one can find the linear entropy of polarization , @xmath17 this corresponds to a completely mixed polarization state in contrast to a completely polarized state with @xmath18 .
we know , the state of polarization represented by a poincare sphere can be completely described by @xmath19 where @xmath20 s and @xmath21 s are the pauli matrices and normalized stokes parameters respectively .
the trace of square of this density matrix is given by @xmath22 where @xmath23 is the degree of polarization which is measured as the magnitude of the stokes vector @xmath24 .
using eq . [ 5 ] and eq . [ 4 ] one can relate @xmath23 to the linear entropy , @xmath25 thus for a maximally non - separable state of polarization and oam , for which @xmath26 , the degree of polarization should be zero .
one can easily determine the dop experimentally by measuring the stokes parameters @xcite .
another characteristic of the non - separable state is the contexuality . for a separable state , measurement on one degree of freedom does nt affect the measurement outcome of the other
however , in the case of a non - separable state , measurement outcome in one degree of freedom will depend on the context of measurement in the other . in our experiment the oam state of the beam varies according to the projections to different polarization states due to their non - separability .
consider a general polarization state defined as @xmath27 where @xmath28 and @xmath29 are the euler angles corresponding to the state @xmath30 on the poincar sphere .
projecting @xmath31 given in eq .
[ 1 ] to @xmath32 , we obtain the oam state as @xmath33 this is a pure oam superposition state .
the transverse profile of the beam will correspond to the superposition of two equal and oppositely charged vortices with different relative amplitudes and phase .
therefore , the intensity profile of the beam varies according to the polarization projections defined by @xmath34 and @xmath35 . for demonstration
we take @xmath36 @xmath37 and @xmath38 which correspond to @xmath39 and @xmath40 polarization states . with @xmath41 for projections to different polarization states .
h - horizontal , v- vertical , d - diagonal , a - anit - diagonal , r - rightcircular , l - leftcircular ] figure [ fg.1 ] shows the theoretical intensity distributions corresponding to different polarization projections for @xmath41 .
the projection on h ( v ) polarization gives a vortex of order @xmath42 .
the projections of the state on diagonal ( d ) , anti - diagonal ( a ) , left circular ( l ) and right circular ( r ) gives superposition of two vortices that contain @xmath43 ( in our case @xmath44 ) number of lobes with different orientations .
the number of lobes confirms the order or the azimuthal index of the vortex and the change in their orientation confirms the presence of non - separability in a light beam .
- lens with focal length 15 cm , ccd- charge coupled device ( camera ) , pm - power meter , pbs- polarizing beam splitter ] the experimental set up used to generate the non - separable state and to study its properties is shown in fig .
we have used a diode pumped solid state green laser ( verdi 10 ) with vertical polarization for our study .
the laser beam passes through a half wave plate , whose fast axis is oriented at @xmath45 with the horizontal that changes beam polarization from vertical to diagonal .
then it passes through a polarizing sagnac interferometer containing a spiral phase plate ( spp ) to generate a light beam which is non - separable in polarization and oam .
two orthogonally polarized ( h and v ) counter propagating gaussian beams are converted into optical vortices of orders @xmath46 ( for h ) and @xmath47 ( for v ) by the spp designed for order @xmath44 .
these orthogonally polarized and oppositely charged vortices superpose at the same pbs to form the described non - separable state .
this non - separable state is generated only in the presence of spp otherwise the superposition of two orthogonally polarized gaussian beams results in a diagonally polarized gaussian light beam .
the doughnut shaped non - separable beam forms a random speckle distribution after scattering through the ground glass . a part of the scattered light collected with a lens of focal length 15 cm placed at a distance of 22 cm from the ground glass plate .
the ground glass plate is rotating at @xmath48 930 revolutions per minute to average out the speckles .
the intensity distributions corresponding to the different polarization projections are recorded with an evolution vf color cooled camera ( pixel size @xmath49 m ) kept at the focus of the lens .
the stokes parameters are measured using a quarter wave plate and a polarizer .
we project the beam to horizontal ( h ) , vertical ( v ) , diagonal ( d ) , anti - diagonal ( d ) , right circular ( r ) and left circular ( l ) polarizations and measure the intensity .
the intensity measurements for determining the stokes parameters were performed with an optical power meter ( thorlab ) of sensitivity 1 nw .
one can find out the stokes parameters as @xmath50 where @xmath51 is the total intensity of the beam and @xmath52 is the intensity corresponding to @xmath53-polarization .
we have measured the stokes parameters @xmath54 of coherent and scattered light beams for both separable ( without spp ) and non - separable states ( with spp ) .
we compare the degree of polarization of beams before and after scattering and the results are given in table [ tb.1 ] . from the table , it is clear that the separable light beam is completely polarized ( diagonal ) while the non - separable state is completely unpolarized .
the deviations in degree of polarization may be due to uncertainties in the orientation of the wave plates , small misalignment of the interferometer and the measurement uncertainty of the power meter . however , our experimental findings are very close to theoretical predictions given in section [ sec.2 ] .
.stokes vectors and the degree of polarization corresponding to separable and non - separable states of light before and after scattering . [ cols="^,^,^,^,^,^,^ " , ] we also generate non - maximally entangled states simply by controlling intensities in the two arms of the interferometer .
this can be done easily by rotating the fast axis of the hwp .
then the state becomes @xmath55 plot for coherent and scattered non - separable states of light along with theoretical curve given by eq .
[ 3 ] . ] by varying @xmath56 from 0 to @xmath51 and correspondingly @xmath57 from @xmath51 to 0 , we have generated different states given in eq . [ 2 ] .
note that the total intensity , @xmath58 is always constant . for the state described in eq . [ 2 ] , we can check the mixedness of the subsystem ( here polarization ) by calculating @xmath16 which also indicates the degree of non - separability .
it reduces to a simple analytic expression , @xmath59 line curve in fig .
[ fg.3 ] shows the variation of linear entropy @xmath16 of polarization with the normalized intensity in one arm of the interferometer as given in eq .
the linear entropy becomes zero when @xmath60 or @xmath61 , for which the state become @xmath62 and @xmath63 respectively .
when the two intensities are same ( @xmath64 ) , the state becomes completely non - separable for which @xmath65 .
we measure the stokes parameters and calculate the degree of polarization and linear entropy experimentally corresponding to each value of @xmath56 for coherent and scattered light beams .
the results are shown in fig .
one can clearly see that the @xmath16 _ vs. _ normalized intensity curve for both the coherent and scattered light are in good agreement with the theoretical curve .
the results of polarization measurements given in table [ tb.1 ] and fig .
[ fg.3 ] which confirm the preservation of non - separability in polarization and oam under scattering by the rgg . for different polarization projections .
oam states corresponding to each intensity distribution are also given . ]
figure [ fg.5 ] shows the intensity distributions for a coherent and a scattered light beam with non - separable state projected to the different polarizations .
our results show the similar behavior for both coherent and scattered light beams and are in good agreement with the theoretical images shown in fig .
[ fg.1 ] that again confirm the preservation of non - separability .
we also observe that the amount of scattered light collected by the lens is irrelevant regarding the non - separable properties .
in fact , one can use multiple number of lenses and collimate the scattered light again to form several copies of a partially coherent non - separable beam .
this property can be used in public communication systems .
in conclusion , we have produced a light beam with non - separable polarization and orbital angular momentum states using a simple interferometer and experimentally verified the preservation of the non - separability under scattering through a rotating ground glass .
the polarization measurements and the images of the beam projected to different polarizations show the presence of non - separability for coherent and scattered light . we have also demonstrated the generation of non - maximally non - separable states of light and studied their behavior under scattering by measuring the degree of polarization .
10 url # 1`#1`urlprefixhref # 1#2#2 # 1#1 b. n. simon , s. simon , n. mukunda , f. gori , m. santarsiero , r. borghi , n. mukunda , r. simon , a complete characterization of pre - mueller and mueller matrices in polarization optics , j. opt .
am . a 27 ( 2010 ) 188199 .
e. karimi , j. leach , s. slussarenko , b. piccirillo , l. marrucci , l. chen , w. she , s. franke - arnold , m. j. padgett , e. santamato , spin - orbit hybrid entanglement of photons and quantum contextuality , phys .
a 82 ( 2010 ) 22115 .
v. dambrosio , n. spagnolo , l. del re , s. slussarenko , y. li , l. c. kwek , l. marrucci , s. p. walborn , l. aolita , f. sciarrino , photonic polarization gears for ultra - sensitive angular measurements . , nat
. commun . 4 ( 2013 ) 2432 .
a. kumar , j. banerji , r. p. singh , hanbury brown twiss - type experiments with optical vortices and observation of modulated intensity correlation on scattering from rotating ground glass , phys .
a 86 ( 2012 ) 13825 .
e. karimi , s. slussarenko , b. piccirillo , l. marrucci , e. santamato , polarization - controlled evolution of light transverse modes and associated pancharatnam geometric phase in orbital angular momentum , phys . rev . a 81 ( 2010 ) 053813 .
s. slussarenko , v. dambrosio , b. piccirillo , l. marrucci , e. santamato , the polarizing sagnac interferometer : a tool for light orbital angular momentum sorting and spin - orbit photon processing . ,
express 18 ( 2010 ) 2720527216 . | we experimentally show that the non - separability of polarization and orbital angular momentum present in a light beam remains preserved under scattering through a random medium like rotating ground glass .
we verify this by measuring the degree of polarization and observing the intensity distribution of the beam when projected to different polarization states , before as well as after the scattering .
we extend our study to the non - maximally non - separable states also .
non - separability , orbital angular momentum , scattering , linear entropy |
frustrated magnetism in kagome heisenberg antiferromagnetic systems ( khaf ) has been a major subject in condensed matter physics due to their unconventional , exotic ground states which emerge from the interplay between geometrical frustration and low - dimensional quantum effects.gfm2005,gfm2011 in particular , quantum fluctuations for @xmath15 systems are found to be strong among khaf and can lead to interesting behavior like that of a spin liquid.@xcite theoretical studies on ideal @xmath15 isotropic khaf lattice have demonstrated that it has a resonating valence bond ( rvb)-like disordered ground state.@xcite recent numerical studies @xcite have also predicted that its ground state is a spin liquid with a small spin - gap ( @xmath16 ) of @xmath17 to its triplet excited state ( where @xmath18 is the exchange interaction between nearest - neighbor ( @xmath19 ) spins ) .
a limited number of experimental realizations of structurally ideal , @xmath20 khaf have been found , which include zn and mg - herberthsmithite , kapellasite , and haydeeite.lists=1/2kagome,e .
fak 2012 among these , zn - herberthsmithite zncu@xmath0(oh)@xmath21cl@xmath22 is the best example to explain isotropic @xmath15 khaf behavior .
experimental studies on zn - herberthsmithite have shown that there is no magnetic ordering down to @xmath23 k , which implies that it has a disordered ground state.@xcite but an unambiguous proof for the existence of a spin - gap from an experimental point of view is lacking .
the natural ion exchange of cu and zn or mg is inevitable in these systems , which is probably the main obstacle to detect the intrinsic nature of these @xmath15 khaf .
on the other hand , there are a few anisotropic khaf systems such as volborthite cu@xmath0v@xmath22o@xmath1(oh)@xmath22.2h@xmath22o @xcite and vesignieite bacu@xmath0v@xmath22o@xmath24(oh)@xmath22 @xcite , which do not have much atomic site - disorder , in which the kagome layers are built by two nearest neighbor ( @xmath19 ) exchange couplings .
these have also been studied to understand the real ground state physics of a kagome system . despite the presence of significant anisotropy
, these systems show interesting spin dynamics.@xcite there is another kind of anisotropic kagome lattice in rb@xmath22cu@xmath0snf@xmath25@xcite where the kagome layer is formed by four exchange couplings and has a spin - gap of @xmath26 k. the pinwheel valence bond solid ( vbs ) excitations are realized in this anisotropic khaf system.k .
matan nature phys2010 * * * * theoretical predictions also suggest that when the kagome layer is perturbed with additional exchange couplings such as next - nearest neighbor @xmath27 and diagonal couplings , the system is driven to a novel ordered state from the disordered state of isotropic kagome layer.@xcite these predictions * * * * prompted us to look for newer @xmath20 khaf systems to explore interesting physics associated with the presence of anisotropy and additional couplings .
herein , we introduce a new anisotropic magnetic system , pbcu@xmath0teo@xmath28 @xcite which has @xmath15 staircase ( buckled ) kagome layers built by cu and o atoms ( see fig .
@xmath29 ) .
this system has kagome layers similar to those of a@xmath0v@xmath22o@xmath24 ( a=@xmath30-cu , ni , co ) .
cu3v2o8,ni3co3synthesis detailed studies have been done on single crystals of @xmath31 and @xmath32 staircase kagome systems ni@xmath0v@xmath22o@xmath24 and co@xmath0v@xmath22o@xmath33 the ni@xmath0v@xmath22o@xmath24 system has four competing magnetic transitions ( two incommensurate and two commensurate ) below @xmath34 k , one of which is a multiferroic transition at @xmath35 k. glawesprl2004,g lawesprl2005 however , co@xmath0v@xmath22o@xmath24 does not show multiferroicity , which might be due to its high spin state with low quantum fluctuations .
less work has been done on the @xmath15 analog @xmath30-cu@xmath0v@xmath22o@xmath36 probably due to non availability of single crystals .
we report here the magnetic properties of pbcu@xmath0teo@xmath1 .
the magnetic susceptibility @xmath2 data indicate that the dominant exchange interactions between cu@xmath5 ions are antiferromagnetic ( af ) with a curie - weiss temperature ( @xmath37 ) of about @xmath38 k. the @xmath2 deviates from curie - weiss behavior below @xmath39 k. we observed a total of three magnetic anomalies ( @xmath7 k , @xmath8 k , and @xmath40 k ) in the @xmath2 data of a single crystal .
the magnetic anomalies at @xmath12 and @xmath13 were evident only when the applied magnetic field @xmath41 was parallel to the crystallographic @xmath42-axis , whereas the one at @xmath43 could be observed only for @xmath44 .
the anomaly at @xmath13 is first order in nature and is field sensitive . on the other hand , heat capacity data in zero field ( for both polycrystalline and single crystal samples )
showed anomalies of @xmath12 and @xmath43 only .
the first - order transition at @xmath13 could not be observed by us in heat capacity data .
we suggest that this might be due to the small latent heat involved in this transition .
our electronic structure calculations in conjunction with our experimental findings suggest the presence of various competing magnetic couplings , in addition to @xmath19 , in pbcu@xmath0teo@xmath1 which in turn cause a deviation from the superficially anticipated regular kagome geometry leading to long - range order ( lro ) .
polycrystalline samples of pbcu@xmath0teo@xmath1 were prepared by conventional solid - state reaction method using pbo , cuo , and te precursors .
the stoichiometric amount of chemicals were ground thoroughly and fired at 750 @xmath45c for @xmath46 days with three intermediate grindings . to obtain single crystals , a mixture of pbcu@xmath0teo@xmath1 and nacl / kcl flux in the mass ratio @xmath47
was charged in an alumina crucible , fired at @xmath48 @xmath49c for @xmath50 hrs and then cooled slowly to @xmath51 @xmath49c with a cooling rate of @xmath29 @xmath45c per hour .
single crystals were extracted after washing the flux with hot water .
the x - ray diffraction ( xrd ) data were collected using @xmath52 advance ( bruker ) on the single crystal and on ( polycrystalline ) powder obtained from crushing the crystals .
magnetic and heat - capacity measurements were performed on polycrystalline and single crystal samples using a physical property measurement system ( ppms ) from quantum design . for heat - capacity , we employed the thermal relaxation method with two time constants ( @xmath53 model ) to fit the heat - capacity data .
the single crystal used for the magnetization measurements is of mass @xmath54 mg and approximate dimensions @xmath55 mm x @xmath56 mm x @xmath57 mm as shown in fig . @xmath58 .
the single phase nature of the polycrystalline sample was confirmed by comparing the measured xrd pattern with the calculated one generated by powdercel@xcite using the initial structural parameters of the orthorhombic space group _ pnma _ ( space group no .
@xmath59 ) given by b. wedel , _ _ et al.__@xcite rietveld refinement of the xrd pattern was done using the fullprof suite software@xcite ( as shown in fig .
@xmath60 ) .
the lattice parameters obtained from the refinement are @xmath61 , @xmath62 , and @xmath63 .
the obtained atomic positions and occupancies are summarized in table @xmath29 .
these structural parameters are in good agreement with the previously published values.@xcite a rectangular crystal was placed on a glass slide with the top face ( see fig .
@xmath54 ) parallel to the plane of the slide and xrd was performed .
it resulted in the appearance of only @xmath64 peaks as shown in fig .
@xmath58 ) .
this indicates that the top crystal surface ( as indicated in fig .
2 ) is perpendicular to the @xmath65 $ ] direction or @xmath66plane .
this means that the @xmath67axis lies in the as - grown plane of the crystal .
note that crystallographic @xmath67axis is perpendicular to the staircase kagome layers , as shown in fig .
we have further taken the laue pattern of the crystal , and found that the crystallographic @xmath67axis is along the length of the crystal .
the other crystallographic axes ( @xmath69 and @xmath70 ) were also identified and make roughly @xmath71 with the top plane of the crystal , as shown in the inset of fig . @xmath58 .
the staircase kagome planes in the pbcu@xmath0teo@xmath1@xcite structure are formed by cu atoms ( cu1 and cu2 ) as shown in fig . @xmath29 .
the cu@xmath72 ions are coupled via o@xmath73 ions and form buckled layers in the @xmath74-planes , as shown in fig . @xmath68 . according to the cu - cu bond - lengths and the cu - o - cu bond - angles
, the staircase kagome plane possibly has four different nearest neighbor @xmath75 exchange constants @xmath76 , @xmath77 , @xmath78 and @xmath79 ( see fig .
@xmath68 ) . in this staircase kagome plane ,
cu1 atoms form linear chains with the couplings alternating between @xmath77 and @xmath80 .
these cu1 chains are connected with each other via cu2 atoms with couplings @xmath76 and @xmath81 to form a @xmath4 network .
other possible magnetic coupling paths are also shown in fig .
@xmath82 , the hopping integrals for which have been calculated by us .
these are @xmath83 @xmath84 @xmath85 and @xmath86 which are in addition to the @xmath19 hoppings @xmath87 @xmath88 @xmath89 and @xmath90 ( corresponding to @xmath76 , @xmath77 , @xmath81 and @xmath79 , respectively ) .
the kagome planes are separated by pb and te atoms with a small inter - planar distance of about @xmath91 , which suggests that significant three - dimensional couplings between the layers might be present .
the magnetization @xmath92 was measured as a function of temperature @xmath93 .
the dependencies of the magnetic susceptibility @xmath94 in the range from @xmath95 k to @xmath96 k for the polycrystalline and single crystal samples of pbcu@xmath97teo@xmath1 are shown in fig .
magnetic measurements were performed on the @xmath54 mg single crystal ( shown in fig .
@xmath58 ) for each of the crystal direction @xmath42 , @xmath99 and @xmath100 respectively in an applied magnetic field @xmath41 .
the @xmath2 data follow a curie - weiss behavior in high-@xmath93 region . even in the paramagnetic region
, a significant anisotropy is seen between the field orientation along @xmath42 and the other two perpendicular directions ( @xmath69 and @xmath101 .
this means that the intra - kagome ( @xmath66plane ) anisotropy ( difference between @xmath102 and @xmath103 axis ) is small compared to the inter - kagome anisotropy .
the @xmath2 data of the polycrystalline sample lie in between the single crystal data for @xmath104 and @xmath105 but for @xmath106 k they are close to the single crystal data of @xmath107 while below @xmath10 k they are closer to the single crystal data for @xmath108 .
the temperature independent magnetic susceptibility @xmath109 was estimated from polycrystalline data at high@xmath110 from @xmath111 k to @xmath48 k ( data not shown here ) .
we subtracted @xmath112 @xmath113/mol cu from our measured data and then plotted @xmath114 vs @xmath93 in fig . @xmath115 . the core diamagnetic susceptibility ( @xmath116 ) of pbcu@xmath0teo@xmath1
is calculated to be @xmath117 @xmath113/mol cu from @xmath116 of individual ions.@xcite the van vleck paramagnetic susceptibility is then estimated to be @xmath118 @xmath113/mol cu , which is comparable with other cuprates.@xcite the @xmath2 data follow the curie - weiss ( cw ) law @xmath119 in the @xmath120range @xmath10 - @xmath96 k where @xmath121 is the curie constant and @xmath37 is the curie - weiss temperature .
the obtained parameters from the single crystal data are given in the table ii .
the @xmath121 value corresponds to an effective magnetic moment ( @xmath122 of about @xmath54 @xmath123/cu , which is slightly larger than the spin - only value for @xmath15 which is @xmath124 @xmath123 .
the negative @xmath37 indicates that the dominant couplings are antiferromagnetic in nature .
there is a deviation from the curie - weiss law below @xmath10 k , which indicates that it is not a simple paramagnet at low@xmath110 .
this deviation might be a result of the competitive exchange couplings present in the quasi-@xmath4 staircase kagome planes .
the @xmath2 data of polycrystalline samples exhibit a total of three magnetic anomalies at @xmath7 k , @xmath8 k , and @xmath9 k , respectively .
the ordering temperature ( @xmath12 ) is relatively smaller than @xmath37 and the temperature below which a deviation from cw behavior is seen ( @xmath10 k ) , which indicates that the system is moderately frustrated .
these three magnetic anomalies ( @xmath12 , @xmath125 and @xmath13 ) are also seen in the single crystal data , but with significant anisotropy .
all of the anomalies are not evident for each field direction .
for instance , the magnetic anomaly @xmath12 is evident only when @xmath104 , and the anomaly at @xmath43 is seen only for @xmath41 @xmath126 . on the other hand for @xmath41 @xmath127 , no clear anomalies were observed .
the anomaly at @xmath13 is prominent for @xmath104 ( see fig .
@xmath128 ) .
although , the appearance of anomalies in the @xmath2 data depends on the field direction , the transitions are not field driven because the two anomalies ( @xmath12 and @xmath43 ) appear in zero - field heat capacity data as will be discussed in a later section .
similar @xmath129direction dependent anomalies were also reported in a multiferroic , spiral magnet fevo@xmath130.@xcite the susceptibility @xmath131 ( @xmath2 for @xmath132 @xmath127 ) increases below the transition while @xmath133 and @xmath134 decrease below the transition .
this suggests that @xmath102axis might be the hard - axis of magnetization .
we have also performed dielectric constant and electric polarization versus @xmath93 down to @xmath54 k on the polycrystalline samples , neither anomaly in the dielectric constant nor electric polarization are observed below the transition temperatures . in order to obtain insight about the transition at @xmath13 , we have measured @xmath92 of a collection of carefully oriented crystals ( @xmath135 ) , while cooling and warming in different @xmath136 of @xmath29 koe , @xmath46 koe , @xmath137 koe , @xmath138 koe , respectively , as shown in fig . @xmath91 .
the transition at @xmath13 is sharp and looks different from those of at @xmath12 and @xmath125 .
there is a difference in the position of the peak for the zero - field - cooled warming ( zfcw ) data and the field - cooled cooling ( fcc ) data , while no difference is observed between zfc warming and fc warming ( fcw ) data .
this kind of thermal hysteresis is a characteristic feature of a first - order phase transition.@xcite similar sharp , first - order anomalies are also observed at @xmath139 k in ni@xmath0v@xmath22o@xmath24@xcite and @xmath140 k in co@xmath0v@xmath22o@xmath24 @xcite staircase kagome systems .
moreover , the observed peak position moves to higher temperatures with increasing @xmath41 and this variation ( summarized in table iii ) suggests that the observed first - order transition is ferromagnetic in nature .
the @xmath13 peak also broadens and the difference between the position ( in temperature ) of the warming and cooling peak decreases as @xmath41 increases . this kind of field - induced broadening is also a characteristic feature of the first - order transition.@xcite finally , the @xmath13 peak is suppressed in a field of about @xmath26 koe , however @xmath12 and @xmath43 peaks remain , more or less , unaffected by magnetic fields upto @xmath141 koe and do not exhibit any thermal hysteresis like observed for @xmath13 ( data not shown here ) .
the @xmath120dependent heat - capacity @xmath142 of the polycrystalline pbcu@xmath0teo@xmath143 samples in the range from @xmath54 to @xmath111 k is presented in fig . @xmath46 .
the @xmath3 data were measured using the ppms by the thermal relaxation method . for a magnetic insulator ,
one expects both lattice and magnetic contributions to the heat - capacity .
we used the debye model@xcite to obtain the lattice part in the following manner .
we fit the @xmath3 data in the @xmath120range from @xmath144 k to @xmath111 k with a linear combination of two debye integrals as given below @xmath145 here @xmath146 is the number of atoms per formula unit , @xmath147 is a debye temperature , and @xmath148 ( @xmath149 ) . the fitting yields @xmath150 @xmath151 k@xmath152 @xmath153 and @xmath154 k. the fitting curve was extrapolated down to @xmath54 k and this was then subtracted from the measured @xmath155 magnetic heat - capacity @xmath156 was thus obtained ( see inset of fig .
@xmath46 ) .
sharp anomalies are observed at@xmath157 k and @xmath158 k , which agree well with the anomalies found at @xmath12 and @xmath43 in @xmath2 .
we observed the presence of @xmath156 upto @xmath144 k which is well above @xmath12 , which signifies the presence of magnetic correlations above @xmath12 and this behavior is also consistent with the magnetic data .
however , we could not detect any transition at @xmath13 in the data .
in fact , detailed @xmath159 measurements were done around @xmath160 k with a large number of points .
the raw data of the sample temperature as a function of time at each of these temperatures showed a good fit to the @xmath53 model . in the case of a first - order transition , it has been documented that deviations from the simple @xmath53 behavior occur for @xmath142 data at the transition temperature.@xcite in light of the missing supporting evidence for @xmath13 from the @xmath3 measurement , one may suspect that the existence of @xmath13 from susceptibility data may not be intrinsic . on the other hand , the inability to detect similar kind of first - order phase transition has also been argued in the literature.@xcite some other technique like the adiabatic method or a better pulse - sequence design@xcite may be needed for measuring small latent heat or entropy change transitions . although our heat capacity measurements do not support the existence of the @xmath13 phase transition , we must stress that the @xmath13 anomaly has been consistently observed in the magnetization measurements for different batches of powder samples as well as single crystals , i.e. , the possibility of contribution from external magnetic impurity phase due to different preparation conditions or contamination can be ruled out .
more sensitive heat capacity measurement are being planned to clarify this issue .
overall we conclude that the intrinsic origin of this transition is an open question .
the entropy change @xmath161 calculated from the magnetic heat - capacity is about @xmath162 j / mol k cu , which is in good agreement with the expected @xmath163 ( @xmath164 j / mol k ) for @xmath165 systems .
the @xmath166 value at transition @xmath12 is found to be @xmath167 j / mol k cu , which is @xmath168 of the total entropy and the rest of the entropy is released in the paramagnetic region well above @xmath12 , suggesting the presence of frustrated correlations .
the observation of significant entropy well above ordering temperature is generally observed in frustrated spin systems.@xcite additionally , nearly all the entropy change has taken place by the time one approaches ( with decreasing temperature ) the @xmath13 transition .
this is the reason i presume that @xmath169 is a weak transition such as a slight canting from the af ordered state to produce weak fm moment .
this might explain the difficulty in observing the transition in heat capacity . in order to study the electronic structure of pbcu@xmath0teo@xmath28
we have carried out first principles density functional theory ( dft ) calculations within the local - density approximation ( lda ) by employing the stuttgart tblmto-47 code based on the linear muffin - tin orbital ( lmto ) method in the atomic sphere approximation ( asa).@xcite the basis set for the self - consistent electronic structure calculations for pbcu@xmath0teo@xmath1 in tb - lmto asa includes pb ( s , p ) , cu ( s , p , d ) , te ( s , p ) and o ( s , p ) and the rest are downfolded .
a ( 4 , 8 , 4 ) @xmath170-mesh has been used for self - consistency .
all the @xmath170-space integrations were performed using the tetrahedron method . in order to ascertain the accuracy of our asa calculations
we also performed the electronic structure calculation using projected augmented wave ( paw ) @xcite method encoded in the vienna @xmath171-initio simulation package(vasp).@xcite the density of states calculated by these two different approaches is found to agree well with each other . in order to extract various hopping integrals between the cu atoms ,
we have employed the n@xmath172 order muffin tin orbital ( nmto ) downfolding method.@xcite the non - spin - polarized band structure for pbcu@xmath0teo@xmath1 is displayed in fig .
the bands are plotted along the various high symmetry points of the brillouin zone corresponding to the orthorhombic lattice .
all the energies are measured with respect to the fermi level of the compound .
the characteristic feature of the non - spin - polarized band structure displayed in fig .
@xmath140 is an isolated set of twelve bands crossing the fermi level and these bands are predominantly derived from the antibonding linear combination of cu - d@xmath173 and o - p states .
these twelve bands are well separated from the low - lying o - p and other cu -d valence bands .
this isolated cu - d@xmath173 twelve band complex is responsible for the low - energy physics of this compound .
@xmath174 shows the non - spin - polarized density of states ( dos ) of pbcu@xmath0teo@xmath1 . as expected , there is strong hybridization between the o - p and cu - d states .
the occupied pb-6s states lie below the fermi level and empty pb - p and te - s states are above the fermi level .
we have employed the nmto downfolding method to map our lda results to a low - energy orthogonal tightbinding hamiltonian by integrating out the high energy degrees of freedom from the all - orbital lda calculation .
the fourier transformation of the downfolded hamiltonian @xmath175 gives the various hopping integrals @xmath176 between the cu atoms .
these hopping integrals will determine the dominant exchange paths for the system .
for the present compound we have kept only the cu - d@xmath173 orbital for each cu atom in the unit cell in the basis and integrated out all the rest to derive the low - energy model hamiltonian .
we show the downfolded bands in fig .
@xmath140 in comparison to the full lda band structure and the agreement between the two is found to be excellent .
the hopping integrals ( @xmath177 10 mev ) obtained from the nmto downfolding method are listed in table iv and v and the exchange paths are indicated in fig . @xmath82 and @xmath178 . the strongest hopping in the kagome plane is @xmath179 between the corner - sharing cu1 and cu2 ions .
this hopping is primarily mediated by o2 situated in the basal plane of the cu1 octahedron forming a strong @xmath180 antibond with the cu1 @xmath181 orbital . the @xmath19 hopping @xmath182 between the edge sharing cu1 octahedron ( see fig .
@xmath183 ) and cu2 tetrahedron ( see fig .
@xmath184 ) , is found to be much smaller than @xmath179 . in order to obtain further insights , we have plotted the wannier function of cu2 @xmath181 orbital corresponding to the @xmath179 and @xmath182 hoppings ( see fig .
@xmath185 ) .
the plot reveals that the cu2 @xmath181 orbital forms strong @xmath186 antibonds with the neighboring oxygens .
we can see that the weight of wannier function at cu1 site which is at a distance @xmath187 is large compared to the weight at cu1 site which is at a distance @xmath188 indicating @xmath189 hopping will be stronger in comparison to @xmath190 .
the bond angle of cu1-o2-cu2 corresponding to the @xmath89 hopping is @xmath191 , while the bond angle of cu1-o1-cu2 is @xmath192 for the @xmath182 hopping path .
the strength of antiferromagnetic interactions are strongly dependent on the angle between the bonds .
when the angle is close to @xmath193 , the af super - exchange process is suppressed due to the orthogonality of cu 3d and o 2p orbitals .
so the exchange coupling along the path @xmath182 is expected to be ferromagnetic . for @xmath90 hopping ,
two cu1 octahedra are corner - shared with each other and the bond - angle of cu1-o3-cu1 is @xmath194 while for the @xmath195 hopping path , the two cu1 octahedra share an edge with each other and the bond angle of cu1-o2-cu1 is @xmath196 . as argued earlier , the @xmath90 hopping dominates over the @xmath195 hopping path due to the corner sharing topology of the cu1 octahedra .
the second strongest hopping is @xmath197 between the cu2 atoms .
this cu2-cu2 hopping primarily proceeds via the oxygens . as a result
, the strength of the cu - o4-o4-cu spin exchange is primarily governed by the o - o distance and in this case o - o distance is 2.57 ( smaller than the van der waals distance ) . moreover , the bond - angle of cu - o4-o4 is about @xmath198 which makes this hopping stronger than some of the @xmath19 hoppings ( @xmath199 @xmath200 and @xmath90 ) .
the hopping paths perpendicular to the kagome plane are listed in table v. since the distances between the cu ions in two different kagome plane is small , it is expected that there is substantial amount of hoppings between the kagome planes .
the strongest interaction between the kagome plane is @xmath201 .
this interaction is mediated via te following the path cu - o - te - o - cu .
our electronic structure calculations reveal that intra - kagome and inter - kagome exchange interactions are long ranged . in general ,
an ideal 2d kagome system with only uniform @xmath19 couplings does not order even at zero@xmath202 .
however , spin wave theory@xcite predicts that the presence of @xmath203 @xmath19 ( @xmath204 ) and @xmath205 @xmath19 ( @xmath206 ) couplings in the kagome plane can drive the system to an ordered ground state .
when @xmath207 , @xmath208 nel ordered state is favored , while the @xmath209 nel state occurs in the case of @xmath210 a rich theoretical magnetic phase diagram has also been built based on the type ( antiferro and ferromagnetic ) and the relative strength ( @xmath211 ) of interactions in the @xmath212 kagome system.@xcite since the kagome plane of pbcu@xmath0teo@xmath1 appear to have additional couplings ( @xmath203 and @xmath213 @xmath19 ) , they might support ordered magnetic states as mentioned above . in addition , there are also substantial inter - planar couplings found in this system .
these additional intra - kagome and inter - kagome couplings might be the reason to have different af transitions in this system .
a detailed neutron diffraction measurements and analysis will however be needed to clarify the origin and nature of these transitions .
based on the structural details , one forms the impression that pbcu@xmath0teo@xmath143 might have quasi-@xmath4 staircase ( buckled ) kagome layers .
however , our magnetic measurements on polycrystalline and single crystal samples show the onset of lro at about @xmath214 k with a change in the spin - order around @xmath158 k. this indicates the existence of significant three - dimensional couplings .
the @xmath2 is found to obey the simple curie - weiss law above @xmath10 k with a curie constant as expected for a paramagnetic system .
however , below @xmath215 k there is a deviation from the curie - weiss behavior though it is well above @xmath216 , which suggests the presence of frustrated spin correlations well above @xmath216 .
this behavior is also consistent with @xmath3 which infers the presence of magnetic entropy well above the magnetic transitions .
we performed electronic structure calculations to determine the relative values and importance of various exchange paths .
our results suggest the presence of various intra - kagome ( @xmath27 and diagonal ) and inter - kagome exchange hopping paths , which then must be responsible for the onset of lro in this system .
detailed neutron diffraction measurements will be helpful to understand more about the nature of spin orderings and magnetic phase diagram of this system .
the authors fcc and bkr acknowledge the support from national science council of taiwan under project number nsc-100 - 2119- m-002 - 021 .
jc thanks csir india ( grant no . 09/080(0615)/2008-emr
-1 ) for research fellowship . avm and rk thank the department of science and technology and csir , india for financial support .
work at snu was supported by national creative research initiative ( 2010 - 0018300 ) .
e. kermarrec , p. mendels , f. bert , r. h. colman , a. s. wills , p. strobel , p. bonville , a. hillier , and a. amato , phy .
b. * 84 * , 100401(r ) ( 2011 ) , o. janson , j. richter , and h. rosner , phys .
* 101 * , 106403 ( 2008 ) .
b. fk , e. kermarrec , l. messio , b. bernu , c. lhuillier , f. bert , p. mendels , b. koteswararao , f. bouquet , j. ollivier , a. d. hillier , a. amato , r. h. colman , a. s. wills , phy .
lett . * 109 * , 037208 ( 2012 ) .
r. nirmala , ya .
mudryk , v. k. pecharsky , and k. a. gschneidner , jr . ,
b * 76 * , 104417 ( 2007 ) .
d. p. rojas , j. i. espeso , j. rodrguez fernndez , j. c. gmez sal , j. sanchez marcos , and h. mller , phys .
b * 80 * , 184413 ( 2009 ) .
y. chen , j. w. lynn , q. huang , f. m. woodward , t. yildirim , g. lawes , a. p. ramirez , n. rogado , r. j. cava , a. aharony , o. entin - wohlman , and a. b. harris , phys .
b * 74 * , 014430 ( 2006 ) , y. yasui , y. kobayashi , m. soda , t. moyoshi , m. sato , n. igawa , and k. kakurai , j. phys .
. jpn . * 76 * , 034706 ( 2007 ) .
j. c. lashley , m. f. hundley , a. migliori , j. l. sarrao , p. g. pagliuso , t.w .
darling , m. jaimea , j. c. cooley , w. l. hults , l. morales , d. j. thoma , j. l. smith , j. boerio - goates , b. f. woodfield , g. r. stewart , r. a. fisher , n. e. phillips , cryogenics * 43 * , 369 ( 2003 ) , h. suzuki , a. inaba , and c. meingast , cryogenics * 50 * , 693 ( 2010 )
. fig .
@xmath29 ( a ) the crystal structure of pbcu@xmath0teo@xmath1 viewed along the @xmath217direction is shown .
( b ) projection of staircase ( buckled ) kagome planes in the @xmath66plane formed by cu and o atoms with various bond - angles @xcite . the possible nearest neighbor ( @xmath19 ) exchange couplings between cu1 and cu2 atoms in the kagome layer
are denoted by @xmath218and @xmath79 .
the environments of the cu1 octahedron , the cu2 tetrahedron and their coupling are shown in ( c ) , ( d ) , and ( e ) .
lda calculations suggest various hopping paths for ( f ) intra - kagome plane and ( g ) inter - kagome planes indicated by @xmath219 , and @xmath220 respectively .
( color online ) fig .
@xmath221 ( a ) powder x - ray diffraction pattern of pbcu@xmath0teo@xmath1 at room temperature .
the open circles indicate the experimental data , while the calculated xrd with residual factors of @xmath222 @xmath223 , and @xmath224 is shown as a solid red line .
the bragg peak positions are indicated by short - vertical marks and the bottom green line represents the difference between the experimental and calculated data .
( b ) the x - ray diffraction pattern when the x - rays are incident at the `` top '' surface of the crystal in the horizontal plane ( the incident and the diffracted beam form the vertical plane ) .
the crystallographic axes as determined from laue diffraction are also illustrated .
( color online ) fig .
@xmath225 ( a ) magnetic susceptibilities ( @xmath226 of a polycrystalline and a single crystal sample are plotted as a function of @xmath227 the inset shows an expanded view of the low - temperature data .
( b ) ( @xmath228 is plotted as a function of @xmath93 .
the data are fitted to the curie - weiss law in the @xmath120range @xmath10 k - @xmath96 k. ( color online ) fig .
@xmath91 magnetic susceptibility of pbcu@xmath0teo@xmath1 in the orientation of @xmath104 for different magnetic fields ( a ) @xmath29 koe , ( b ) @xmath46 koe , ( c ) @xmath137 koe and ( d ) @xmath138 koe is plotted .
the black solid line indicates zero - field - cooled ( zfc ) warming data , while the red and blue solid lines are the data related to field - cooled ( fc ) cooling and fc warming , respectively .
the red and black arrow marks indicate the position of magnetic anomaly in the cooling and warming data , respectively .
( color online ) fig . @xmath46
the @xmath142 data ( open circles ) are plotted as a function of @xmath93 .
the red solid line is a fit to the equation ( see text ) in the @xmath120range from @xmath144 k to @xmath229 k and the red dashed line is the extrapolation of the fit .
inset shows magnetic heat capacity ( left @xmath230axis ) and entropy ( right @xmath230axis ) versus @xmath93 .
( color online ) | we have synthesized polycrystalline and single crystal samples of pbcu@xmath0teo@xmath1 and studied its properties via magnetic susceptibility @xmath2 and heat - capacity @xmath3 measurements and also electronic structure calculations . whereas the crystal structure is suggestive of the presence of a quasi-@xmath4 network of cu@xmath5 @xmath6 buckled staircase kagome layers , the @xmath2 data show magnetic anisotropy and three magnetic anomalies at temperatures , @xmath7 k , @xmath8 k , @xmath9 k , respectively .
the @xmath2 data follow the curie - weiss law above @xmath10 k and a curie - weiss temperature @xmath11 k is obtained .
the data deviate from the simple curie - weiss law below @xmath10 k , which is well above @xmath12 , suggesting the presence of competing magnetic interactions .
the magnetic anomaly at @xmath13 appears to be of first - order from magnetization measurements , although our heat - capacity @xmath3 results do not display any anomaly at @xmath14 the hopping integrals obtained from our electronic structure calculations suggest the presence of significant intra - kagome ( next - nearest neighbor and diagonal ) and inter - kagome couplings .
these couplings take the pbcu@xmath0teo@xmath1 system away from a disordered ground state and lead to long - range order , in contrast to what might be expected for an ideal ( isotropic ) @xmath4 kagome system . |
the symmetry and the dimensionality are important factors to determine the universality class of critical phenomena .
the @xmath2 symmetry is the simplest among the continuous symmetry , and statistical models with the @xmath2 symmetry has been studied intensively . a natural question
then would be the effect of the symmetry breaking from the continuous @xmath2 to the discrete @xmath0 .
a simple spin model with @xmath0 symmetry is the @xmath3-state clock model with a hamiltonian @xmath4 where @xmath5 runs over nearest neighbors , and @xmath6 takes integral multiples of @xmath7 .
the standard xy model with @xmath2 symmetry is defined by the hamiltonian of the same form ; the only difference is that @xmath8 takes continuous values .
the @xmath0 symmetry is fundamentally different from @xmath2 because of its discrete nature . on the other hand , for large @xmath3
, it is natural to expect the @xmath0 symmetry to have similar effects to that of the @xmath2 symmetry .
understanding these two apparently contradictory aspects is an interesting problem .
besides the theoretical motivation , there are some possible experimental realizations of the effective @xmath0 symmetry . for example
, the stacked triangular antiferromagnetic ising ( sti ) model with effective @xmath1 symmetry may correspond to materials such as@xcite csmni@xmath9 . in two dimensions ,
the phase diagram of the @xmath0 model is well understood@xcite in the framework of the renormalization group ( rg ) . for @xmath10 ,
there is an intermediate phase between the low - temperature ordered phase with the spontaneously broken @xmath0 symmetry and the high - temperature disordered phase .
the intermediate phase is @xmath2 symmetric and corresponds to the low - temperature phase of the xy model . on three dimensional ( 3d ) case ,
blankschtein et al.@xcite in 1984 proposed an rg picture of the @xmath1 models , to discuss the sti model .
they suggested that the transition between the ordered and disordered phases belongs to the ( 3d ) xy universality class , and that the ordered phase reflects the symmetry breaking to @xmath1 in a large enough system .
it means that there is no finite region of rotationally symmetric phase which is similar to the ordered phase of the xy model .
unfortunately , their paper is apparently not widely known in the related fields .
it might be partly because their discussion was very brief and not quite clear .
in fact , there has been a long - standing controversy on the the three - state antiferromagnetic potts ( afp ) model on a simple cubic lattice , defined by the hamiltonian @xmath11 where @xmath12 and @xmath5 runs over nearest neighbor pairs on a simple cubic lattice .
the order parameter of this model is not evident .
however , previous studies revealed that the low - temperature ordered phase , which is called as broken sublattice symmetry ( bss ) phase@xcite , corresponds to a spontaneous breaking of the @xmath1 symmetry . thus the effective symmetry of this model may be regarded@xcite as @xmath1 , although it is not apparent in the model .
it is now widely accepted that there is a phase transition with critical exponents characterized by the 3d xy universality class@xcite , at temperature @xmath13 ( we set the boltzmann constant @xmath14 . ) on the other hand , according to numerical calculations , there appears to be an intermediate phase below @xmath15 and above the low - temperature phase .
while there have been various proposals@xcite for the intermediate region , most reliable numerical results at present indicates that the intermediate region appears to be rotationally symmetric phase which is similar to the ordered phase of the 3d xy model@xcite .
however , the `` transition '' between the intermediate region and the low - temperature phase is not well understood .
according to the suggestion in ref . , the intermediate `` phase '' would be rather a crossover to the low - temperature massive phase .
on the other hand , there has been a claim of an intermediate phase@xcite also in the @xmath16-state clock ( 6cl ) model , which has the manifest @xmath1 symmetry . in a recent detailed numerical study
, miyashita@xcite found that the intermediate region appears to have a rotationally symmetric character , as found in the afp model .
however , through a careful examination of the system size dependence , he concluded that it is just a crossover to the massive low - temperature phase , and that the rotationally symmetric xy phase does not exist in the thermodynamic limit .
his conclusion is consistent with the suggestion in ref . .
in this article , based on the rg picture , we derive a scaling law of an order parameter which measures the effect of symmetry breaking from @xmath2 to @xmath0 .
we demonstrate that the monte carlo results on the afp model in ref .
is consistent with the scaling law , supporting the rg picture with a single phase transition .
since the discussion of the rg picture in ref . was rather brief , it would be worthwhile to present the rg picture here , with some clarifications and more details .
we also make a straightforward extension to general integer @xmath3 from the @xmath17 case .
a generic @xmath0 symmetric model may be mapped , in the long - distance limit , to the following @xmath18-type field theory with the euclidean action @xmath19\ ] ] with the complex field @xmath20 and its conjugate @xmath21 .
the @xmath22-term is the lowest order term in @xmath20 which breaks the symmetry from @xmath2 to @xmath0 .
the phase transition corresponds to the vanishing of ( the renormalized value of ) the parameter @xmath23 .
the temperature @xmath24 in the @xmath0 statistical system roughly corresponds to @xmath23 as @xmath25 where @xmath15 is the critical temperature . in the absence of the symmetry breaking @xmath22
, the transition belongs to the so - called 3d xy universality class .
its stability under the symmetry breaking to @xmath0 is determined by the scaling dimension of @xmath22 at the 3d xy fixed point .
it may be estimated with the standard @xmath26-expansion method .
the lowest order result in @xmath26 can be easily obtained from the operator product expansion ( ope ) coefficients@xcite . as a result
, we obtain the scaling dimension @xmath27 of @xmath22 in @xmath28 dimensions as @xmath29 @xmath27 is defined so that the effective strength of the perturbation @xmath30 at scale @xmath31 is proportional to @xmath32 near the xy fixed point .
the case @xmath33 is actually the special case @xmath34 of the `` cubic anisotropy '' on the 3d @xmath35 fixed point@xcite . extrapolating the @xmath36 result to 3d ( @xmath37 )
, we see that the @xmath0 perturbation is irrelevant at the 3d xy fixed point for @xmath38 .
the threshold @xmath39 is estimated to be @xmath40 in @xmath36 in fact , @xmath41 and @xmath42 corresponds to the 3d ising and 3-state ( ferromagnetic ) potts model , which do not belong to xy universality class .
thus @xmath39 is expected to be at least @xmath40 .
this is consistent with the above result from @xmath36 .
however , extrapolating the lowest order result in @xmath26 to 3d ( @xmath37 ) is not quite reliable ; the true value of @xmath39 might be larger than @xmath40 .
on the other hand , we can make following observation . for @xmath43 , @xmath22 is marginal or irrelevant at the 3d gaussian fixed point ( @xmath44 ) .
thus it is natural to expect them to be irrelevant at the more stable 3d xy fixed point , namely @xmath45 .
in fact , the numerical observation of the 3d xy universality class in 6cl and afp model strongly suggests that @xmath46 is irrelevant at the xy fixed point and hence @xmath45 . in the following ,
we restrict the discussion to the irrelevant case @xmath38 . for the @xmath2 symmetric case @xmath47
, low - temperature phase @xmath48 is renormalized to the low - temperature fixed point .
it describes the massless nambu - goldstone ( ng ) modes on the groundstate with the spontaneously broken @xmath2 symmetry .
let us call the low - temperature fixed point as ng fixed point . in terms of the field theory ,
it is described by the @xmath2 sigma model ( free massless boson field ) @xmath49 where @xmath50 is the angular variable @xmath51 .
namely , only the angular mode @xmath50 remains gapless as a ng boson . in three dimensions , the coupling constant @xmath52 renormalizes proportional to the scale @xmath31 , and goes to infinity in the low - energy limit .
the coupling constant may be absorbed by using the rescaled field @xmath53 so that the action is always written as @xmath54 .
now let us consider effects of the symmetry breaking @xmath22 .
the symmetry breaking term can be written as @xmath55 . using the rescaled field @xmath8 , the total effective action at scale @xmath31 becomes @xmath56 } , \ ] ] where the factor @xmath57 comes from the scale transformation of the integration measure . in the thermodynamic limit
, we should take @xmath58 limit .
physically , it means that the @xmath2 symmetry is spontaneously broken so that the angle is fixed to some value @xmath59 in a single infinite system
. then the taylor expansion of the cosine in @xmath60 becomes valid : @xmath61 } = \sum_{j=0}^{\infty } c_j k^{3-j/2 } \theta^j , \ ] ] where @xmath62 for a nonnegative integer @xmath63 .
the five terms @xmath64 are relevant perturbations . for any value of @xmath59 , some of the coefficients @xmath65 of these relevant terms are non - vanishing .
we therefore conclude that , unlike the 2d case , the @xmath0 perturbation is always relevant for any value of @xmath3 at the ng fixed point . we emphasize that this conclusion is universal in three dimensions and independent of the microscopic model . shortly speaking
, the @xmath0 perturbation gives mass to the pseudo ng boson @xmath8 , which would be massless ng boson in the absence of the perturbation .
in contrast , in two dimensions the coupling constant @xmath52 of the free boson field theory is dimensionless , and the above argument does not apply .
it is related to the absence of a spontaneous breaking of a continuous symmetry .
we now have a global picture of the rg flow as shown in fig .
[ fig : rgflow ] .
the phase transition between the ordered phase and the disordered phase is governed by the xy fixed point .
this means that the critical exponents are identical to those of the xy model .
this is consistent with the numerical results . in the disordered phase above @xmath15
, there will be no essential effect of the @xmath0 perturbation .
however , the nature of the ordered phase is more interesting .
the @xmath0 perturbation @xmath22 is eventually enhanced in the ordered phase below @xmath15 .
it means that all region below @xmath15 belong to the massive phase with the spontaneously broken @xmath0 symmetry .
there is no rotationally symmetric intermediate phase , unlike the 2d case . only a precisely @xmath2 symmetric model with @xmath66
is renormalized to the ng fixed point below @xmath15 , corresponding to the rotationally symmetric low - temperature phase .
an interesting aspect of the rg flow diagram is that the @xmath0 perturbation is irrelevant at the 3d xy fixed point but is relevant at the low - temperature ng fixed point .
this could be related to a nontrivial system size dependence found in a monte carlo renormalization group calculation@xcite . for @xmath24
slightly less than @xmath15 , the symmetry breaking perturbation @xmath22 is renormalized to a small value by the rg flow , and remains small until the rg flow reaches near the ng fixed point .
it means that the mass of the pseudo - ng bosons is suppressed by the fluctuation effect . at a finite scale ( for example in a finite size system ) , the ordered phase near @xmath15 is very similar to the low - temperature phase of the xy model .
this naturally explains the numerical observation of the apparently rotationally symmetric `` phase '' in 6cl or the afp model . for larger @xmath3 ,
the mass is more suppressed , and the low - temperature side of the transition appears to be @xmath2 symmetric until the system size becomes very large . however , for any finite @xmath3 , the low - temperature side of the transition @xmath67 is not truly massless nor @xmath2 symmetric in the thermodynamic limit , as already pointed out .
based on the rg picture , we derive a scaling law on an order parameter @xmath68 which characterizes the symmetry breaking from the @xmath2 to @xmath0 symmetry .
there are various possible definitions of @xmath68 . on the 6cl model ,
miyashita@xcite numerically measured an order parameter @xmath69 which corresponds to the effective barrier height . on the afp model , heilmann , wang and swendsen@xcite studied @xmath70 , which is the fourier transform of the angle distribution density of average spins .
the following consideration apply to the both cases . for large enough @xmath71 and @xmath24 slightly lower than @xmath15
we divide the rg flow to three stages , as shown in fig .
[ fig : scaling ] ( i ) : : the rg flow near the 3d xy fixed point .
the symmetry breaking @xmath22 is irrelevant , and is renormalized proportional to @xmath72 at length scale @xmath31 .
( ii ) : : the rg flow from the neighborhood of the 3d xy fixed point to the ng fixed point . for simplicity , we assume that the symmetry breaking @xmath22 is unchanged in this stage .
( iii ) : : the rg flow near the ng fixed point .
@xmath22 is relevant , giving a mass to the ng boson .
the length scale @xmath73 , at which the crossover from stage ( i ) to ( ii ) occurs , is given by @xmath74 , where @xmath75 is the correlation length exponent of the 3d xy universality class .
thus , at the crossover , @xmath76 this also gives the effective value of the perturbation @xmath22 at the crossover from stage ( ii ) to stage ( iii ) . in the presence of the @xmath0 perturbation
, the spin configuration would be dominated by the ordered regions which are separated by domain walls in a large system .
the free energy costed by the domain walls is proportional to their area , which scales as @xmath77 for the system size @xmath71 .
therefore the effective `` barrier height '' is proportional@xcite to @xmath77 . combining this with eq .
( [ eq : lmdscale ] ) , we conclude that the order parameter is a function of a single scaling variable : @xmath78 where @xmath79 is a constant .
the function @xmath80 is universal , but of course depends on the definition of the @xmath68 . while the scaling by @xmath77 was used in ref . , we find that the temperature dependence of the order parameter is also governed by a scaling .
interestingly , the exponent @xmath81 is completely determined by the 3d xy fixed point .
in the numerical study@xcite of the afp model , they claimed the existence of the intermediate phase , in which the order parameter @xmath70 is very small even for relatively large lattice ( upto @xmath82 ) .
however , we re - analyze their data to demonstrate the scaling relation ( [ eq : orderp ] ) , and hence the validity of the rg picture . in fig .
[ fig : scaling ] , we show the data taken@xcite from fig . 3 of ref . .
we chose @xmath83 to give the best scaling .
the data for various temperature and various system sizes fall remarkably into a single curve as a function of the scaling variable @xmath84 .
this supports the proposed scaling relation ( [ eq : orderp ] ) .
furthermore , if we approximate the effective potential by @xmath85 , the scaling function is given by @xmath86 where @xmath87 is the modified bessel function .
choosing @xmath88 , the scaled data agree with this simple function rather well .
we note that the data appear to deviate from the scaling law for small @xmath89 .
this may be due to the insufficient system size @xmath71 or the relatively large statistical error .
we emphasize that the present scaling relation is a strong evidence of the single phase transition at the temperature @xmath15 .
in contrast , the scaling of the `` spontaneous magnetization '' @xmath90 does not distinguish our picture and the `` intermediate phase '' scenario of ref . . on the other hand , the scaling function @xmath91 for @xmath69 in ref
is linear in @xmath89 by definition .
he indeed found that @xmath69 is scaled by @xmath77 .
however he did not discuss the temperature depedence .
we have attempted to analyze the data in figs . 6 and 7 in ref .
, to find that they are roughly consistent with our scaling law ( [ eq : orderp ] ) with the exponent @xmath92 .
the estimate is difficult because there are only small number of temperature points available in ref . .
according to our picture , the exponent @xmath93 is a universal quantity determined by the 3d xy universality class .
considering the available data , the above results on afp and 6cl models are consistent with the universality hypothesis , although not conclusive .
it would be interesting to obtain more numerical data on these models to check our scaling law ( [ eq : orderp ] ) .
the exponent @xmath75 has been determined as @xmath94 for the 3d xy universality class@xcite . combining with the above estimates of @xmath93 , @xmath95 is estimated as about @xmath16 .
unfortunately , the irrelevant eigenvalue @xmath96 has not been much discussed in the literature .
the lowest order result ( [ eq : yn1 ] ) in the @xmath26-expansion gives @xmath97 , which is not quite consistent with the numerical estimate .
however , it is perhaps not surprising to obtain an inaccurate result in the lowest order of the @xmath26-expansion
. it would be interesting to carry out the calculation to higher orders in @xmath26 , or to estimate @xmath96 by other means .
in this article , we clarified a rg picture of phase structure of 3d @xmath0 symmetric models , which was introduced earlier by blankschtein et al.@xcite .
there is no finite region of intermediate phase with a ( spontaneously broken ) @xmath2 symmetry , but only a crossover to a massive phase where the discreteness of @xmath0 is relevant .
based on the rg picture , we have derived a scaling law of the order parameter in the 3d @xmath0 models the existing monte carlo data@xcite on the afp model , which was used to claim the intermediate phase , was shown to be consistent with the scaling law .
thus we conclude that the rg picture is valid on the afp model , and there is only one transition at @xmath13 with the 3d xy universality class .
we would like to make a few final remarks .
firstly , we note that the rg argument used in the present article does not contradict to the transition of other than xy universality class , because only the local stability of the xy fixed point was discussed .
it is possible that a lattice model with @xmath0 symmetry is renormalized to another ( unknown ) rg fixed point .
actually , it appears somewhat controversial whether the transition of the sti model belong to the xy universality class@xcite . on the other hand
, the available numerical results strongly supports that the 6cl and afp models at the critical temperature are renormalized into the xy fixed point .
once the transition is known to be xy universality class , the rg picture and the scaling law discussed in this article should apply to the ordered phase , for the temperature slightly below the critical point . secondly , as discussed in ref .
, the `` bare '' value of @xmath22 ( at a small length scale ) may have opposite sign in some circumstances .
namely , the minima and maxima of the potential of @xmath50 are swapped . in such a case
, the ordered phase may correspond to the permutationally symmetric sublattice ( pss ) phase proposed in ref .
for the afp model , or the incompletely ordered phase ( iop ) proposed in ref . for the 6cl model .
in the vicinity of the critical point , the temperature dependence of the bare @xmath22 is not essential because the leading dependence on the temperature is determined by the critical effect , as shown in eq .
( [ eq : orderp ] ) .
however , it may be important in a wider temperature range . in particular , if the bare @xmath22 changes sign at some temperature @xmath98 lower than @xmath15 , we would have a transition@xcite at @xmath98
such a transition would be controlled by the ng fixed point .
the existing numerical data indicates that there is no such phase transition in the standard afp model on the simple cubic lattice or in the standard 6cl model .
however , such a transition may be possible in some modified models@xcite .
in fact , blankschtein et al.@xcite argued it to exist in the sti model .
i would like to thank hikaru kawamura , macoto kikuchi , ryo kishi , seiji miyashita and yohtaro ueno for useful discussions .
this work is supported in part by grant - in - aid for scientific research from the ministry of education , culture and science of japan .
y. ajiro , t. inami and h. kadowaki , j. phys .
. jpn . * 59 * , 4142 ( 1990 ) ; h. kadowaki , t. ianmi , y. ajiro and y. endoh , j. phys .
. jpn . * 60 * , 1708 ( 1991 ) .
d. nelson , in phase transitions and critical phenomena , vol . 7 ed .
by c. domb and j. l. lebowitz , academic press , london .
d. blankschtein , m. ma , a. n. berker , g. s. grest and c. m. soukoulis , phys .
b * 29 * , 5250 ( 1984 ) .
j. r. banavar , g. g. grest and d. jasnow , phys .
b * 25 * , 4639 ( 1982 ) .
i. ono , prog .
. suppl . * 87 * , 102 ( 1986 ) .
a. rosengren and s. lapinskas , phys .
lett . * 71 * , 165 ( 1993 ) .
y. ueno , g. sun and i. ono , j. phys .
. jpn . * 58 * , 1162 ( 1989 ) .
wang , r. h. swendsen and r. koteck , phys .
b * 42 * , 2465 ( 1990 ) . y. okabe and m. kikuchi , in _ computational physics as a new frontier in condensed matter research _ , ed . by h. takayama et al . , the physical society of japan ( 1995 ) m. kolesik and m. suzuki , physica a * 216 * , 469 ( 1995 ) .
r. k. heilmann , j .- s .
wang , r. h. swendsen , phys .
b * 53 * , 2210 ( 1996 ) .
r. kishi , master thesis , tokyo institute of technology ( 1999 ) .
y. ueno and k. mitsubo , phys .
b * 43 * , 8654 ( 1991 ) .
s. miyashita , j. phys .
* 66 * , 3411 ( 1997 ) .
j. cardy , _ scaling and renormalization in statistical physics _ , cambridge university press ( 1996 ) .
m. kikuchi and y. okabe , unpublished .
the data were taken from the postscript file in the e - print version of ref .
, cond - mat/9509140 .
the postscript file contains useful `` raw '' data , thereby eliminating the errors in digitizing the figure ( or the need to bother the authors asking the old data . )
j. zinn - justin , _ quantum field theory and critical phenomena ( 3rd ed . ) _ , oxford university press ( 1996 ) .
m. l. plumer and a. mailhot , phys .
b * 52 * , 1411 ( 1995 ) .
b. d. gaulin , a. bunker and c. kallin , phys .
b * 52 * , 1415 ( 1995 ) .
y. ueno and k. kasono , phys .
b * 48 * , 16471 ( 1993 ) .
s. lapinskas and a. rosengren , phys .
* 81 * , 1302 ( 1998 ) . | based on a renormalization - group picture of @xmath0 symmetric models in three dimensions , we derive a scaling law for the @xmath0 order parameter in the ordered phase . an existing monte carlo calculation on the three - state antiferromagnetic potts model , which has the effective @xmath1 symmetry , is shown to be consistent with the proposed scaling law .
it strongly supports the renormalization - group picture that there is a single massive ordered phase , although an apparently rotationally symmetric region in the intermediate temperature was observed numerically . |
stellar processing is the main evolutionary process affecting the cosmos .
nucleosynthesis and mass return to the interstellar medium modify the chemical composition of the universe and the fraction of mass bound in stars .
star formation dominates the present - day radiation field , and most of it happens in normal galaxies ( e.g. kim & sanders 1998 ; driver 1999 ) .
relatively little , however , is known about star formation on the scale of a galaxy , including its drivers and inhibitors , and the role of interactions ( e.g. volker 2000 ; dopita et al .
2002 ; hameed & young 2003 ) . the @xmath2 key project on the interstellar medium of normal galaxies ( helou et al .
1996 ; dale et al .
2000 ) aimed to better understand the large - scale physics of the interstellar medium through an array of mid- and far - infrared data on a diverse sample of 69 nearby star - forming galaxies ( e.g. dale et al .
1999 , dale et al .
2000 , helou et al .
2000 , hunter et al .
2001 , malhotra et al .
2001 , contursi et al .
2002 , lu et al .
2003 ) . in this contribution
we present and discuss near - infrared integral field spectroscopy for several of these objects . for optical depths @xmath3 ,
the brightest emission lines that probe the star - forming interstellar medium are found in the near - infrared .
vibrationally excited emission at 2.122 arises by collisional excitation from shocks or radiative excitation in intense ultraviolet radiation fields ( e.g. goldader et al .
1997 ) , or possibly from reprocessed x - ray illumination ( maloney , hollenbach , & tielens 1996 ) .
on the other hand , the lines at 1.257 or 1.644 probe supernova - shocked gas ( the final stage of processing ) and hard power - law / x - ray environments , both of which can release iron atoms and ions through interstellar dust grain sputtering and evaporation ; recent @xmath4 work on m 82 and ngc 253 by alonso - herrero et al .
( 2003 ) indicates that as much as 70% of the flux ultimately derives from supernovae , and only 68% from regions .
in contrast to , hydrogen lines like ( 1.282 ) and ( 2.166 ) directly trace the gas ionized by young massive stars .
thus , an -to - hydrogen line ratio can be interpreted as an indicator of the maturity of the local star formation episode , as reflected in the ratio of supernova - shocked gas to molecular gas feeding star formation .
moreover , the ratios / and ( 2.122)/ discriminate between shock excitation from supernova remnants or hard x - ray heating from power - law sources ( larkin et al .
1998 ) , and are essentially unaffected by reddening .
in addition to studying these diagnostics , we use the near - infrared line and continuum images to explore the maturity and spatial progression in the star formation . coupled with imaging ,
the near - infrared hydrogen line fluxes are used to estimate the optical depth and reconstruct the intrinsic emission , total ionizing flux , and other properties of the local interstellar medium .
data at other wavelengths help to further probe the interstellar medium .
for example , new optical spectroscopy and archival mid - infrared imaging allow us to investigate trends with metallicity and infrared colors
. the various results will point to important physical differences between nuclear and extranuclear/ regions in star - forming galaxies .
the targets ( tables [ tab : sample ] and [ tab : obs ] and figures [ fig:2mass1][fig:2mass5 ] ) derive from the @xmath2 key project on the interstellar medium of normal galaxies .
the key project collected infrared data for 69 star - forming galaxies : carbon , oxygen , and nitrogen fine - structure line fluxes between 50 and 160 ( malhotra et al .
2001 ) , mid - infrared spectra between 3 and 12 ( lu et al .
2003 ) , and mid - infrared maps at 7 and 15 ( dale et al .
in addition , the following ancillary ground - based data have been obtained for a large portion of the sample : broadband @xmath5 and narrowband imaging , long - slit optical spectroscopy , and literature co fluxes .
the key project sample had the original following criteria : @xmath6 ) @xmath7 jy ; a published redshift ; and no agn or seyfert classification on ned ( at that time ) .
the sample explores the full range of morphology , luminosity ( @xmath8 from less than @xmath9 to as large as @xmath10 ) , infrared - to - blue ratio ( 0.05 to 50 ) and infrared colors ( see 2 of dale et al .
2001 for a complete description of the sample selection ) .
the subset of the key project sample selected for this particular ground - based follow - up project was chosen according to three criteria .
the subset should : contain targets of high 7 surface brightness ( to ensure detectable targets in the near - infrared ) ; span a range of properties like metallicity , infrared - to - blue ratio , mid - infrared color , morphology , etc . ; and be constrained in redshift such that multiple galaxies could be observed with the same grating tilts , to minimize observational overheads .
ic 10 is classified as a magellanic - type dwarf irregular galaxy and has been extensively studied in the optical ( e.g. lequeux et al .
1994 ; hunter 2001 ) .
the galaxy appears to exhibit ongoing star formation at a rate of @xmath11 ( thronson et al .
1990 ) and has the highest global surface density of wolf - rayet stars in the local group ( massey & armandroff 1995 ) .
the emission spans some seven times the optical extent of the galaxy , and wilcots & miller ( 1998 ) suggest that ic 10 is still forming through accretion of the surrounding material . ngc 693 is a non - magellanic irregular system with average infrared - to - optical and infrared colors .
this galaxy s fairly high global ratio is presumably indicative of a system dominated by diffuse cirrus emission ( a low ratio would indicate intense star formation ; dale et al . 2000 ) .
this impression is supported by the optical spectrum ( integrated along the major axis ) , which shows strong stellar absorption features in the blue ( e.g. calcium h and k , and higher - order balmer lines ) . ugc 2855 is a late - type spiral with prominent arms that are quite striking in the mid - infrared ( dale et al . 2000 ) .
httemeister , aalto , & wall ( 1999 ) argue that this galaxy lacks evidence for strong shocks and is in a pre - burst stage .
their claim is based on the @xmath12 kpc - long continuous molecular bar , which is relatively undisturbed and rotating in a solid - body fashion .
moreover , the / ratio is low and the emission along the bar is weak .
our optical spectrum of a representative region in ugc 2855 shows very strong nebular emission lines superposed on a weak and relatively featureless underlying continuum .
ugc 2866 is a companion galaxy .
the post - starburst ngc 1569 is famous for its two brightest compact super - star clusters ( `` a '' and `` b '' ) . in this study
we have targeted two of the galaxy s regions .
the southeastern and northwestern regions respectively correspond to regions 7 and 2 of waller ( 1991 ) ; the latter has also been designated as `` region c '' by greve et al .
( 1996 ) and `` gmc 3 '' by taylor et al .
ngc 1569 has unusually strong ionic line intensities relative to measures of the dust and photodissociation region emission , in both the mid- and far - infrared .
it stands out as extreme among the general key project sample in the ratios of [ ] 88/far - infrared , [ ] 88/ [ ] 158 , and [ ] 88/ [ ] 63 , consistent with the idea that the galaxy is powered by vigorous ongoing star formation ( malhotra et al .
the extreme mid - infrared colors point to an intense radiation field ( dale et al . 2000 ) .
the @xmath13 isophot spectra from 24@xmath024 apertures centered on the southeastern and northwestern regions ( but overlapping with the super - star clusters ) are dominated by extraordinarily strong ionic emission , in this case the [ ] 10.5 line , relative to the underlying dust continuum emission .
the optical spectra of regions in this galaxy display extremely strong [ ] 4959 , 5007 emission due to the high electron temperature @xmath14 , a result of the low abundance of oxygen , the primary coolant .
however , since the infrared lines ( see above ) are not sensitive to @xmath14 , the strong line - to - continuum contrast contrast in the infrared must have a different origin , possibly a deficiency of interstellar dust emitting at these wavelengths due to the extremely intense ultraviolet radiation field ( see lu et al .
2003 ) .
ngc 2388 is a spiral galaxy with fairly unremarkable infrared properties .
we did not obtain an optical spectrum of ngc 2388 .
ngc 4418 has the most unusual infrared properties in our sample .
roche et al .
( 1986 ) interpret the deep 9.7 spectral trough in its spectrum as a silicate absorption feature , and infer a high extinction , @xmath15 , towards the nuclear source .
the galaxy has an extremely small ratio and an extremely large ratio , indicative of an intense interstellar radiation field ( dale et al .
its mid - infrared spectrum lacks the typical polycyclic aromatic hydrocarbon features ( lu et al .
ngc 4418 also exhibits an unusually low /far - infrared ratio ( an upper limit , in fact ) ; the interpretation favored by malhotra et al .
( 2001 ) is that interstellar dust grains are highly positively charged in this active galaxy , leading to less efficient heating by photo - ejected electrons from dust grains .
the optical spectrum of ngc 4418 is dominated by spectral features from late - type stars .
the only emission line evident is a weak emission peak near 6620 which may be a component of a complete h@xmath16 profile that includes a small absorption dip at the galaxy s systemic velocity . due to the large extinction towards the infrared - active nuclear region of this galaxy ,
the optical spectrum does not probe this region , but instead simply presents a spectrum characteristic of the older stellar population of the galaxy s disk .
ngc 6946 is a well - studied , nearly face - on galaxy with classic spiral structure and a starbursting core ( engelbracht et al .
schlegel , blair , & fesen ( 2000 ) find 14 pointlike x - ray sources in ngc 6946 , the brightest of which is unremarkable in the mid - infrared ( compare with the mid - infrared maps in dale et al .
the region we targeted to the ne shows up in the @xmath17 hri image as a faint arc @xmath1805 in length , whereas the nucleus is associated with a point source with the third highest x - ray luminosity in the galaxy ( @xmath19 j ) .
ngc 7218 is a late - type spiral with typical infrared properties .
the galaxy has a short bar oriented roughly perpendicular to the major axis .
the optical emission lines show a level of star - forming activity intermediate between ngc 693 and ngc 7771 on the one hand , and ugc 2855 and ngc 1569 on the other .
ngc 7771 has a prominent circumnuclear star - forming ring of major axis diameter 6 - 7 with @xmath2010 radio - bright clumps and @xmath2010 near - infrared - bright clumps ( smith et al .
1999 ; reunanen et al .
the radio and near - infrared sources in the ring are offset from each other , lending to the idea that there are multiple star - forming populations of different ages . a possible interaction with the nearby dwarf ngc 7770
may be related to much of this activity .
smith et al .
( 1999 ) find that the radio regions are @xmath204 myr old , whereas the near - infrared spots are slightly older , some @xmath2010 myr .
reunanen et al .
( 2000 ) present and line maps and find that the starburst ring exhibits small equivalent widths ( of order unity dex ) and their interpretation is an instantaneous starburst that occurred 6 - 7 myr ago .
the spectroscopic observations were carried out between september 2000 and november 2001 ( table [ tab : obs ] ) using the palomar integral field spectrograph ( pifs ; murphy , matthews , & soifer 1999 ) on the palomar 200-inch telescope .
pifs provides a 54@xmath096 field of view using eight separate 067@xmath096 slits to feed two independent spectrographs within the same liquid n@xmath21-cooled dewar .
each slit is four pixels wide and 58 pixels long ( 0167 pixel@xmath22 ) .
the high resolution mode ( @xmath23 ) was used , providing a usable wavelength range of 0.030.04 and a velocity resolution of @xmath24 near the observed wavelengths . the standard observational procedure consisted of a set of four five - minute on - off integrations .
thus 20 minutes were spent integrating on source , and an equal amount of time off source .
the only exceptions to this procedure involved ugc 2855 and ngc 693 , for which we observed 40 minutes and 45 minutes on source , respectively . for all observations ,
the 54@xmath096 field of view was oriented with the long axis at a position angle of 90 .
a positional dithering pattern was employed for the sequence of observations to recover seeing - limited spatial sampling in the cross - slit direction .
this dithering is the reason the resultant maps ( 57@xmath0100 ) are slightly larger than the pifs field of view .
spectral calibration lamp data were taken immediately after each set of four dithered integrations .
corrections for atmospheric opacity and spectral flat - fielding are derived from observations of f8-g9 main sequence stars .
an estimate of the near - infrared psf was periodically obtained , with each measurement comprised of four dithered exposures of a nearby field star using the pifs imaging mode ( see table [ tab : obs ] ) .
the atmospheric conditions for each set of observations are listed in table [ tab : obs ] . since about half of the observations were taken in non - photometric conditions , for uniformity all the observations
are flux - calibrated using 2mass @xmath25 data , via fluxes from cutouts matched to the pifs field of view .
this technique is uncertain .
first , near - infrared broadband fluxes are contaminated at some level by nebular line emission . in very young starbursts
/ blue compact galaxies , for example , the line contribution can be as large as @xmath185 - 10% of the total near - infrared flux ( e.g. figure 3 of krger , fritze - v .
alvensleben , & loose 1995 ; vanzi et al
however , except for ic 10 , ngc 1569 nw , and ngc 6946 ne , none of the equivalent widths for our sample indicate broadband flux contributions larger than 2% .
second , the 2mass and pifs aperture positionings may not exactly coincide .
third , depending on the location of the emission within the pifs field of view , 2mass flux dilution can be a factor since the 2mass images are @xmath26 in resolution whereas the pifs data typically are sub - arcsecond .
based on a comparison of 2mass flux calibration and the calibration we obtained from standard stars on the photometric nights , we estimate near - infrared fluxes in this work are uncertain at the @xmath27% level .
figures [ fig : pifs1][fig : pifs4 ] present line emission contour maps superposed on the continuum images at the same wavelength / grating setting , whereas figures [ fig : spectra1 ] and [ fig : spectra2 ] display the integrated spectra .
a few of the spectral images exhibit strong residuals due to non - uniform illumination during the spectral flat - field procedure .
images of the galaxies in this sample were obtained in several observing runs .
ngc 2388 was observed with the 1.07 m hall telescope at lowell observatory , and the rest of the sample was observed with the perkins 1.8 m telescope at lowell observatory .
ic 10 and ngc 1569 were observed in november 1992 with a ti 800@xmath0800 ccd on loan to lowell observatory from the us naval observatory in conjunction with the ohio state university fabry - perot that was used simply as a 3:1 focal reducer .
the resulting scale was 0488 pixel@xmath22 .
ngc 2388 was observed in december 1997 with a site 2048@xmath02048 ccd , binned 4@xmath04 .
the pixel scale was 062 pixel@xmath22 .
the rest of the observations with the perkins telescope were made in october
december 1996 and used an 800@xmath0800 ti ccd provided to lowell observatory by the national science foundation .
this detector was coupled to a 4:1 focal reducer to yield a pixel scale of 0433 pixel@xmath22 .
ngc 6946 was observed with the kpno 0.9 m telescope in may 1993 , and the data were graciously provided by a. ferguson ( ferguson et al .
the continuum was subtracted using an image taken with a narrowband filter centered at 6409 and width 88 ( a. ferguson , private communication ) .
january 200 kpno 2.1 m data of ngc 1569 were kindly provided by c. martin to enhance our analysis ( martin , kobulnicky , & heckman 2002 ) .
all observations at lowell observatory used a series of redshifted filters with fwhm of 2732 , and an off - band filter centered at 6440 with a fwhm of 95 .
the off - band image was shifted , scaled , and subtracted from the image to remove the stellar continuum and leave only nebular emission .
the images were reduced using the over - scan strip to remove the electronic pedestal and dome flat - field images to remove pixel - to - pixel variations .
multiple images were obtained and combined to remove cosmic rays .
the emission was calibrated using the known flux of ngc 604 ( d. hunter , unpublished spectrophotometry ) and spectrophotometric standard stars with minimal absorption features ( stone 1977 ; oke & gunn 1983 ) .
these objects agreed with each other to 4% and the calibrations remained constant to 8% for a given filter over many observing runs .
the flux has been corrected for contamination and the position of the line within the bandpass , shifted for the outside temperature .
we assume / ( + ) @xmath28 ( kennicutt 1992 ; jansen et al .
2000 ) . to account for the temperature - dependent shifting of the filter bandpass
, we assume the bandpass shifts blueward by 0.18 for every degree celsius colder than 20 c. fluxes for the pifs fields - of - view are extracted from the images .
the overall absolute calibration of these fluxes is estimated at @xmath29% to account for pifs aperture misalignment , unknown observing systematics , and the uncertainties in contamination , which exhibits a 20% dispersion in star - forming galaxies ( kennicutt & kent 1983 ; kennicutt 1992 ; jansen et al .
2000 ) .
optical long - slit spectra of six of the galaxies studied in this paper , ngc 693 , ngc 1569 , ngc 4418 , ngc 7218 , ngc 7771 , and ugc 2855 , were obtained with the large cassegrain spectrograph ( lcs ) on the mcdonald observatory harlan j. smith reflector during observing runs in september 1996 , march 1997 , and september 1997 .
these data are a subset of a larger body of optical spectra of galaxies in the _ iso _ key project , which will be presented and discussed in more detail in a future paper ( dinerstein et al .
, in preparation ) .
the 90@xmath30-long slit of the lcs was opened to a width of 15 , yielding a spectral resolving power of @xmath18800 on the texas instruments 800@xmath0800-pixel ccd detector .
separate setups were used to cover the blue ( @xmath18 3680@xmath315060 ) and red ( @xmath18 4800@xmath317450 ) spectral regions ; both set - ups included h@xmath32 and the [ ] 4959 , 5007 lines .
both blue and red data were obtained for all of the galaxies except ngc 1569 and ugc 2855 , for which only the red spectra were taken .
integration times are given in table [ tab : optical_spectroscopy ] .
standard iraf procedures were used to reduce the data , extract and calibrate apertures corresponding to selected spatial regions , and measure line strengths .
the physical regions sampled by the optical spectra vary from galaxy to galaxy , and do not always correspond to the apertures used for the near - infrared spectroscopy .
the optical slit was usually rotated to the position angle of the galaxy s major axis , but in some cases we then offset onto emission peaks in the images described in [ sec : halpha ] .
for ngc 693 , ngc 4418 , and ngc 7771 , a single spectrum representing the integrated light of the disk was extracted , while for ngc 1569 , ngc 7218 , and ugc 2855 we extracted spectra from smaller apertures centered on bright disk regions .
the primary quantities we derived from the optical spectra are tabulated in table [ tab : optical_spectroscopy ] : @xmath33 , the logarithmic extinction at h@xmath32 as derived from the ratio of h@xmath16 to h@xmath32 ( which we used to correct the optical line ratios for reddening ) ; the measured h@xmath32 flux in the extraction aperture ( uncorrected for extinction ) ; and several bright - line indices used as empirical metallicity indicators .
these include : r@xmath34 = ( [ ] 4959 , 5007 @xmath35 [ ] 3727)/h@xmath32 ; r@xmath36 = ( [ ] 4959 , 5007)/h@xmath32 ; o@xmath37 = ( [ ] 4959 , 5007)/ [ ] 3727 ; and n / o = [ ] 6584/ [ ] 3727 .
the [ ] 6717 , 6731 intensity ratios generally indicate low values for the electron density @xmath38 , @xmath39 ) . for the three galaxies for which we obtained blue data , ngc 693 , ngc 7218 , and ngc 7771 , we used the r@xmath34 calibration of edmunds & pagel ( 1984 ) to obtain the ( o / h ) abundances given in table [ tab : sample ] .
if we had used the calibrations of mcgaugh ( 1991 , 1994 ) for intermediate ionization parameters , the inferred ( o / h ) values would be systematically lower by @xmath310.1 dex , but the galaxies would retain the same order and differences in abundance values . note the contrast between ngc 693 and ngc 7771 , which have similar line fluxes and line - to - continuum contrast ( in integrated light ) but strikingly different ionic line ratios ( table [ tab : optical_spectroscopy ] ) , which indicate that ( o / h ) and ( n / o ) are about [email protected] dex higher in ngc 7771 than in ngc 693 .
these two galaxies give an excellent illlustration of both the excitation - metallicity effect which is the basis of the bright - line method , and the well - known correlation between metallicity and total mass or luminosity of disk galaxies ( see garnett 2002 for a recent discussion ) .
ngc 7218 is intermediate in abundance between ngc 693 and ngc 7771 .
since we have only red spectra for ugc 2855 , we assume that it has the same ( o / h ) as ngc 693 based on the similarity of their r@xmath40 indices .
on the other hand , ngc 1569 has extremely strong [ ] lines .
taking r@xmath40 as a lower limit to r@xmath34 , we find an ( o / h ) value of about 8.2 , consistent with the result of kobulnicky & skillman ( 1997 ) .
near - infrared emission lines are securely detected in 84% ( 37/45 ) of the lines and regions targeted ( table [ tab : fluxes ] ) . in general , is the strongest of the observed lines , and ( 2.122 ) the weakest ( no significant emission is observed for any region in the sample ) .
this is also true for the extinction - corrected fluxes ( [ sec : extinction ] ) .
the most notable exception to this rule is provided by ngc 4418 , which has a surprisingly large flux , brighter in fact than the other three near - infrared lines observed combined .
the equivalent widths are generally smaller for the nuclei than for the regions .
for example , the uncertainty - weighted averages for and are @xmath41 and @xmath42 for nuclei and @xmath43 and @xmath44 for regions .
this is no surprise since the nuclear regions have comparatively much stronger near - infrared continuum emission from older stars .
ho , fillipenko , & sargent ( 1997 ) found a similar result in their study of nuclei and disk regions .
the and equivalent widths are essentially ratios of line emission from regions around young stars to continuum emission from older stars .
if the conditions of either an instantaneous starburst or recent constant star formation are met , then these equivalent widths are excellent age indicators . from a modeler s perspective
, one needs to know the star formation history for at least 1 gyr ; if there are variations in the star formation rate with time , the continuum is difficult to constrain . on the other hand ,
if the recent starburst is dominant and has an age less than 30 myr , its red supergiant population is often so strong that the older population does not significantly affect the inferred starburst age .
for example , in the @xmath45 instantaneous starburst model of starburst99 ( figure 51 of leitherer et al . 1999 ) , the peak in the @xmath46 band luminosity as a function of time ( for @xmath47 ) stems from the burst s red supergiant population ; this peak at @xmath48 yr characterizes the starburst .
compare this luminosity to the luminosity generated by a galaxy - wide continuous @xmath49 yr@xmath22 star formation that occurs for , say , @xmath50 yr ( see their figure 52 ) .
considering that only a small portion of this underlying global luminosity is projected to spatially overlap with an individual region , equivalent widths for an region can indeed be relatively unaffected by an older population .
assuming that the equivalent width can be related to the effective star formation age , the ages derived for an instantaneous starburst ( leitherer et al .
1999 ) range from a few million years for the youngest regions to longer than 10 myr for the oldest nuclear regions ( table [ tab : results ] ) .
the regions tend to have more recent starbursts than nuclei .
if nuclei in general are not defined by a single starbursting population , and instead are comprised of star formation episodes that span a range of ages , then it is expected that their inferred ages would be older than an isolated starburst .
interestingly , excluding the unusual system ngc 4418 , there is perhaps a trend in the starburst age with mid - infrared color .
systems with lower , indicating more active star formation , tend to be younger ( see figure [ fig : density ] ) .
a similar trend was previously reported for a study of galaxy circumnuclear regions ( roussel et al .
finally , readers familiar with post - starburst ngc 1569 may be surprised by its apparently young age .
we emphasize that our observations centered on two regions , not the famous super - star clusters .
moreover , the differing spatial distributions in the continuum and line ( [ sec : spatial_distribution ] ) can lead to ill - defined equivalent widths ; we have extracted an average equivalent width , based on the entire fluxes within the pifs field of view . for the nuclear regions where line emission is detected , the continuum and line emission are generally co - spatial ( have similar morphology ) .
on the other hand , there are clear spatial offsets between the line emission from the ionized gas and the continuum emission in the disk regions ( see the maps in figure [ fig : pifs1][fig : pifs4 ] ) .
the similar distributions of the line ( gas ) and continuum emission ( stars ) in the galaxy nuclei may arise naturally , as a result of interstellar material settling to the center of the galaxy s gravitational potential well , confining the peak density of various components to a narrow region .
conversely , the regions do not provide the same kind of gravitational environment , so that the distribution of ionized gas is affected by a number of other factors .
for example , interstellar material can be blown away from a region that is actively forming stars : the map of ngc 1569 exhibits a hole centered on the two bright super starclusters , and the distinct `` s '' shape to the near - infrared line emission maps in ngc 1569 nw is also apparent in at the same location ( martin , kobulnicky , & heckman 2002 ; the emission line contours for ngc 1569 nw correspond to gmc 3 of taylor et al .
1999 ) .
alternatively , there may simply be two or more populations of different ages , as in the smith et al .
( 1999 ) study of the ngc 7771 circumnuclear star - forming ring and the alonso - herrero et al .
( 2003 ) study of starbursts m 82 and ngc 253 . in ngc 7771 ,
smith et al .
( 1999 ) find that the 1.6 kpc diameter ring appears as @xmath2010 distinct sources in both the radio and near - infrared continua .
however , the radio and near - infrared clumps do not overlap , suggesting at least two different formation timescales .
most of the ring is visible in the pifs field of view , and there is evidence for at least three distinct sources in our map .
alonso - herrero et al .
( 2003 ) find a continuum - line spatial mismatch for much of m 82 and ngc 253 in pa@xmath16 and .
in addition , given that the majority of radio supernova remnants and compact sources are not spatially aligned , these authors suggest that there is a young and old population of supernova remnants .
we derived the extinction for each region from the ratios of ( from the narrow - band imaging ) , , and .
the values listed in table [ tab : results ] are derived by least - squares fitting a screen model to the observed line ratios / and / ( the two line ratios are fit simultaneously ) and assuming the li & draine ( 2001 ) extinction curve . uncertainties on the derived extinctions are based on the observational uncertainties and the predicted range in the intrinsic line ratios : /@xmath51 and /@xmath52 for case b recombination ; @xmath53 ; and @xmath54 ( storey & hummer 1995 ) .
a small difference is found between the nuclear and region extinctions .
the averages for the sample are @xmath550.7 mag for the regions and [email protected] mag for the nuclei .
our results are consistent with previous determinations . for example , reunanen et al .
( 2000 ) find @xmath57(int+mw)@xmath58 mag for the central 11@xmath011 of ngc 7771 , and davies , alonso - herrero , & ward ( 1997 ) obtain @xmath59 mag for the more obscured optical nucleus ( their `` region 2 '' ) .
engelbracht et al .
( 1996 ) find that the core of ngc 6946 is extincted by @xmath57(int+mw)@xmath60 mag for a foreground screen model .
the internal extinction for ngc 1569 se and ngc 1569 nw has been estimated by kobulnicky & skillman ( 1997 ) to be @xmath180.2 mag ( their regions c15 and a12 , respectively ) .
the extinction for ngc 4418 is unconstrained by these data , since the fluxes for both and are upper limits .
the range of extinctions found by ho , filippenko , & sargent ( 1997 ) for a sample of over 200 `` nuclei '' is @xmath61 .
the results of kennicutt , keel , & blaha ( 1989 ) for and starburst nuclei are consistent with this result .
this range of extinction for optically - selected samples is a bit smaller than the average nuclear extinction derived for our sample ( @xmath623 mag ) , with ngc 693 being significantly higher than the average .
in fact , in most cases the extinction values are larger than those indicated by our long - slit - derived ratio of h@xmath32/h@xmath16 ( note : for a ratio of total to selective extinction of @xmath63 , @xmath64 ) .
this is to be expected in regions where dust is mixed with gas , since the optical line emission is seen selectively from the regions of lowest extinction .
as kennicutt ( 1998 ) points out , infrared - selected samples reveal much more dramatic nuclear star formation , and @xmath65 band obscuration can easily exceed several magnitudes in the cores of star - forming galaxies ; we are able to probe deeper at near - infrared wavelengths than optical studies , and the more deeply embedded sources simply are more obscured . the number of hydrogen - ionizing photons is derived from an average of the ( extinction - corrected ) and luminosities n(h^o ) [ s^-1]=\ { 7.63 l_br [ 10 ^ 20 w ] + 1.40 l_pa [ 10 ^ 20 w ] } ( leitherer & heckman 1995 ) . if either the or flux is an upper limit , then only the securely detected flux is used .
table [ tab : results ] also expresses this number in terms of the equivalent number of o8.5 v stars , using a conversion from vacca , garmany , & shull ( 1996 ) .
the average projected density of ionizing photons for this sample , computed from a ratio of the number of ionizing photons to the size of the emitting region , spans two orders of magnitude , from @xmath66 = 46.4 $ ] for ngc 7218 to 48.2 for ngc 1569 nw ( the uncertainty on these numbers is 0.2 dex ) .
a strong and understandable correlation is obtained between the mid - infrared color and ionizing photon density : lower , an indicator of more active star formation , corresponds to more intense radiation environments ( figure [ fig : density ] ) .
frster schreiber et al .
( in preparation ) find a similar result for disk and nuclear regions in a sample of nearby galaxies .
perhaps not surprisingly , the one outlier in the observed trend is ngc 4418 , the most unusual target in our sample ( see [ sec : n4418 ] , [ sec : fluxes ] , and [ sec : extinction ] ) .
this object may falsely appear to be lacking in ionizing photons either because of extreme extinction , or because its regions are very young and still rich in dust which absorbs a large fraction of the ionizing photons .
finally , with the exception of the region dubbed `` ngc 6946 ne , '' the regions exhibit the most intense radiation fields and lowest ratios .
the combination of @xmath67 and the flux can place constraints on the dominant excitation mechanism for .
based on the results of black & van dishoeck ( 1987 ) , @xmath68 for fluorescence triggered by ultraviolet radiation ( fluorescence , however , can initially be much more significant compared to its final equilibrium value , e.g. goldshmidt & sternberg 1995 and hollenbach & natta 1995 ) . for our sample ,
the observed ratio is of order @xmath69 , suggesting that the dominant excitation mechanism is shocks rather than fluorescence .
the / line ratio can be a useful diagnostic of the star formation history since emission is thought to arise from supernova activity .
accordingly , is an indicator of past star formation , whereas signifies current star formation ( e.g. moorwood & oliva 1988 ; calzetti 1997 ) . in
a sense / is a ratio of non - thermal to thermal activity ( however , both thermal and non - thermal mechanisms can be attributed to supernovae and their remnants , e.g. sankrit et al .
1998 ; rho et al . 2003 ) . while other lines also trace the current star formation ,
is particularly useful since / is relatively insensitive to extinction .
infrared - bright galaxies that are predominantly powered by star formation should thus have lower / ratios than agn - dominated systems ( e.g. larkin et al . 1998 ) .
( 2.122@xmath70)-to- is another ratio that in principle is unbiased by extinction .
its interpretation , however , is more difficult .
, in contrast to , can be triggered by both thermal and non - thermal processes .
as described by calzetti ( 1997 ) , molecular hydrogen on the surface of molecular clouds can be collisionally excited in gas heated by slow shocks , ultraviolet radiation , and x - ray radiation , or it can result from fluorescence following ultraviolet absorption . empirically , starburst galaxies and regions exhibit lower / ratios ( larkin et al . 1998 ) .
inspection of figure [ fig : ratios ] confirms that these line ratios help to empirically distinguish between galaxies powered by different processes .
data from the literature show there is an approximate trend proceeding from the upper right to the lower left of increasing star formation activity and decreasing agn activity ( or , in the absence of agn activity , decreasing star formation activity occurring in brief , intense and embedded bursts with shocks ) . in fact , all three regions for which we measured both line ratios reside in the lower left , and the one source in our sample suggested to harbor agn activity , ngc 4418 ( roche et al . 1986 ; baan , salzer , & lewinter 1998 ; spoon et al .
2001 ) , is the upper - rightmost data point in our sample .
the nuclear observations for the remainder of our sample lie between these extremes .
it should be noted , however , that systems with low metallicity might be expected to have low iron - to - hydrogen ratios regardless of the source that powers their infrared emission ; the location of ic 10 and ngc 1569 in figure [ fig : ratios ] must reflect in part their low metallicity .
following frster schreiber et al .
( 2001 ) , the gas - phase abundance of fe@xmath1 can be estimated assuming a single - layer model of uniform density and temperature : n(fe^1 + ) n(h ) = l([feii ] ) j(h ) l(h ) j([feii ] ) [ eq : abundance ] where @xmath71 is the observed line luminosity , @xmath72 the number density , and @xmath73 the line emissivity . is the reference h line in this work .
the emissivities are computed in the low density limit @xmath74 ( hamann & ferland 1999 ; quinet , le dourneuf , & zeippen 1996 ) , so we assume collisional de - excitation is unimportant when computing line emissivities . ]
assuming @xmath75 k ( storey & hummer 1995 ; hamann & ferland 1999 ) .
the fe@xmath1 abundances are listed in table [ tab : results ] . for comparison ,
the total gas - phase iron abundance in the orion nebula is estimated to be of order fe / h@xmath76 ( see rubin et al .
1997 and references therein ) .
one particular object exhibits extreme properties in essentially every category .
ngc 4418 is isolated in plots of near - infrared line ratios , ionizing photon density and starburst age . unlike every other source
, it was undetected in and , and is the brightest of the four transitions sampled .
it is possible that ngc 4418 deviates from the overall age - infrared color trend due to a more complicated star formation history .
some claim the galaxy harbors a heavily obscured agn ( roche et al .
1986 ; spoon et al .
2001 ) , while others suggest it is a pre - starburst galaxy .
roussel et al .
( 2003 ) , for example , use new multi - wavelength radio data to argue that galaxies like ngc 4418 may be experiencing the very onset of a starburst after a long time of quiescence , giving them a distinctive infrared emission signature in both continuum and lines .
other studies have shown ngc 4418 to likewise be peculiar ( [ sec : n4418 ] ) . though our data do not have the spectral and spatial resolution to resolve this debate , they similarly point to an object that is not currently powered by star formation .
near - infrared spectroscopy over two spatial dimensions has been obtained for the first time for nuclear and extranuclear star - forming regions in a diverse sample of nine nearby galaxies .
the sample was previously observed with @xmath2 as part of the key project on the interstellar medium of normal galaxies .
mid - infrared and other ancillary data are used in conjunction with continuum and line maps at , , , and wavelengths to explore the physical processes underlying star formation . in this small sample ,
the regions differ from the galaxy nuclei according to nearly every metric .
nuclear observations typically show spatial coincidence of the continuum and line emission , unlike that seen for star - forming extranuclear regions .
in addition , the star - forming episodes characterizing the regions appear to be younger and more intense than in the nuclei .
one hypothesis is that the density of material in nuclear regions supplies continual , on - going episodes of star formation , while extranuclear star - forming regions are more transient and dependent upon random overdensities in the interstellar distribution .
thus if measures of a single star - formation age are made , such as that inferred from equivalent widths , on - going nuclear star formation would naturally lead to longer timescales than that seen for individual starbursting episodes in regions .
moreover , continual nuclear star formation means both the near - infrared continuum and line transitions would be strong at the same locations , even if they represent stellar populations of different ages . alternatively , though there may be cycling in the nuclear star formation history , the strong central density gradient should result in spatially coincident continuum and line emission .
though the regions appear to differ from the nuclei by most measures , this result is perhaps not surprising nor profound , especially given our limited sample of targets .
near - infrared line ratios that distinguish between agn and star - forming activity are consistent with the regions being dominated by star formation whereas the nuclear regions reflect a mixture of star formation and a secondary ionization component ( in the absence of agn activity , the secondary component could be due to embedded and intense starburst - induced shocks ) .
perhaps the nuclei are not exclusively powered by normal star formation , echoing the suggestion of terlevich et al .
( 1992 ) that agn activity is the natural consequence of the final stages of intense nuclear star formation .
ho , filippenko , & sargent ( 1997 ) find that about half ( 42% ) of the bright spirals have nuclear emission spectra with region - like line ratios ; the percentage is a steep function of global morphology , with up to 80% of the later types exhibiting nuclei .
kennicutt , keel , & blaha ( 1989 ) find that approximately `` half of the region and starburst nuclei show evidence for a secondary ionization component , either an active galactic nucleus or large - scale shocks ... [ supporting ] the idea that many spiral nuclei are composite in nature , with a central liner or seyfert - like nucleus surrounded by star - forming regions . ''
consistent with this interpretation , our nuclear data do not exhibit extreme line ratios indicative of pure star formation or agn dominance , but instead suggest a combination of the two processes .
one of the secondary goals of this work was to search for any trends in the sample as a function of metallicity .
no obvious correlations are seen with respect to the oxygen abundances .
however , on a purely empirical basis the regions show lower gas - phase fe@xmath1 abundances than nuclei by an order of magnitude . _
total _ gas - phase fe abundances can not be computed with these data since only the singly - ionized species has been observed .
these data are additionally limited since fe@xmath1 abundances are difficult to interpret .
not only do fe@xmath1 abundances depend on the overall amount of fe produced , fe@xmath1 abundances also depend on the ionizing strength of the interstellar radiation field and how much fe has been depleted onto , or released from , dust grains .
these complications are likely why the abundances do not correspond in any way to the derived metallicities .
nevertheless , there is an interesting contrast in the fe@xmath1/h abundance for the -like regions and the nuclei .
the -like regions show low abundances in comparison to the nuclei .
lack of abundance data from additional ionic species leaves many questions unanswered . for example , do environments with harder radiation fields like regions exhibit relatively higher fe@xmath77 and fe@xmath78 abundances , helping to explain their relatively low fe@xmath1 abundances ?
models of orion nebula data indirectly support this scenario , with ionization fractions of fe@xmath1/fe@xmath77/fe@xmath78=0.0128/0.244/0.744 ( baldwin et al .
1991 ) and 0.0529/0.414/0.533 ( rubin et al . 1991 ) .
to directly compare these fractions in a relatively soft and a relatively hard radiation environment , we utilize the nebula software ( rubin et al .
1991 as updated recently rodrguez & rubin 2003 ) . keeping the metallicity ( orion - like ) and the number of ionizing photons ( 10@xmath79 ) the same from simulation to simulation , we find a relatively small change in the fe@xmath1 abundance but a clear shift to higher fe@xmath78 abundances , mostly at the expense of fe@xmath77 , for harder radiation fields ( see table [ tab : ratios ] ) .
perhaps the most interesting result is that portrayed in the bottom panel of figure [ fig : density ] , the relation between ionizing photon density and mid - infrared color .
relatively intense radiation fields result in lower , likely via two different effects : reduced emission from polycyclic aromatic hydrocarbons and a steepening mid - infrared continuum at 15 ( e.g. dale et al .
1999 ; dale et al . 2001
; helou et al .
intense radiation fields could result from higher photon densities or harder radiation ( roussel et al .
2001 ) . to explore this issue
, we turn to o@xmath37 , the ratio of 5007@xmath354959 to 3727 flux .
is not the best discriminator for radiation hardness for systems with @xmath80 , like ngc 7771 and ngc 6946 ( see figure 8 of kobulnicky , kennicutt , & pizagno 1999 ) . ]
three de - reddened measurements are listed in table [ tab : optical_spectroscopy ] , and four de - reddened values are drawn from elsewhere : logo@xmath81 for ic 10 , the nucleus of ngc 6946 , ngc 1569 nw , and ngc 1569 se , respectively ( lequeux et al .
1979 ; heckman , crane , & balick 1980 ; c. kobulnicky , private communication ) . once again
, the regions behave differently than the nuclei , logo@xmath82 and logo@xmath83 , with the regions exhibiting harder radiation fields .
this o@xmath37 segregation according to environment implies that mid - infrared colors are in part determined by the hardness of the radiation field .
we owe gratitude to several who unselfishly contributed : j. larkin for near - infrared line ratio data ; a. ferguson for an image of ngc 6946 ; c. kobulnicky , c. martin and b. buckalew for ngc 1569 images ; m. regan for preliminary investigations of co data ; d.g . hummer and h.w.w . spoon for their program that extracts intensity ratios of hydrogenic recombination lines ; and c. leitherer and a. bianchini for enlightening discussions .
we also would like to acknowledge the palomar staff and b.t .
soifer who helped make the near - infrared observations possible .
suggestions by the referee improved this work .
this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation .
this research has made use of the nasa / ipac extragalactic database which is operated by jpl / caltech , under contract with nasa .
iraf is distributed by the national optical astronomy observatories , which are operated by the association of universities for research in astronomy , inc .
, under cooperative agreement with the national science foundation .
dum alonso - herrero , a. , rieke , g.h . ,
rieke , m.j . , &
kelly , d. 2003 , , 125 , 1210 baan , w.a .
, salzer , j.j . , & r.d .
lewinter 1998 , , 509 , 633 baldwin , j.a . ,
ferland , g.j . ,
martin , p.g . ,
corbin , m.r .
, cota , s.a . , peterson , b.m . , & slettebak , a. 1991 , , 374 , 580 black , j.h . & van dishoeck , e.f .
1987 , , 322 , 412 calzetti , d. 1997 , , 113 , 162 contursi , a. , kaufman , m. , helou , g. , hollenbach , d. , brahuer , j. , stacey , g. , dale , d.a . ,
malhotra , s. , rubio , m. , rubin , r. , & lord , s. 2002 , , 124 , 751 dale , d.a . ,
helou , g. , silbermann , n. , contursi , a. , malhotra , s. , & rubin , r. 1999 , , 118 , 2055 dale , d.a . ,
silbermann , n.a . ,
helou , g. , contursi , a. et al .
2000 , , 120 , 583 dale , d.a . ,
helou , g. , contursi , a. , silbermann , n. , & kolhatkar , s. 2001 , , 549 , 215 dale , d.a . , & helou , g. 2002 , , 159 , 576 davies , r.i . , alonso - herrero , a. , & ward , m.j .
1997 , , 291 , 557 dopita , m.a .
, pereira , m. , kewley , l.j . , & capaccioli , m. 2002 , , 143 , 47 driver , s.p .
1999 , , 526 , l69 edmunds , m.g . & pagel , b.e.j .
1984 , , 211 , 507 engelbracht , c. w. , rieke , m.j .
, rieke , g.h . , & latter , w.b .
1996 , , 467 , 227 frster schreiber , n.m . ,
genzel , r. , lutz , d. , kunze , d. , & sternberg , a. 2001 , , 552 , 544 ferguson , a.m.n . ,
wyse , r.f.g . ,
gallagher , j.s . , & hunter , d.a .
1998 , , 506 , l19 garnett , d.r .
2002 , , 581 , 1019 goldader , j.d . ,
joseph , r.d . ,
doyon , r. , & sanders , d.b .
1997 , , 474 104 goldschmidt , o. & sternberg , a. 1995 , , 439 , 256 greve , a. , becker , r. , johansson , l.e.b .
, & mckeith , c.d .
1996 , , 312 , 391 hamann , f. & ferland , g. 1999 , , 37 , 487 hameed , s. & young , l.m .
2003 , , 125 , 3005 heckman , t.m . , crane , p.c .
, & balick , b. 1980 , , 40 , 295 helou , g. , malhotra , s. , beichman , c.a . , dinerstein , h. , hollenbach , d.j . , hunter , d.a .
, lo , k.y . , lord , s.d . , lu , n.y . ,
rubin , r.h . , stacey , g.j . ,
thronson , jr . , h.a . , & werner , m.w . 1996 , , 315 , l157 helou , g. , lu , n.y . ,
werner , m.w . ,
malhotra , s. , & silbermann , n.a .
2000 , , 532 , l21 helou , g. , malhotra , s. , hollenbach , d. , dale , d.a . , & contursi , a. 2001 , , 548 , l73 ho , l. , filippenko , a.v . , &
sargent , w.l .
1997 , , 487 , 568 hollenbach , d. & natta , a. 1995 , , 455 , 133 hunter , d. , kaufman , m. , hollenbach , d. et al .
2001 , , 553 , 121 hunter , d. 2001 , , 559 , 225 httemeister , s. , aalto , s. , & wall , w.f .
1999 , , 346 , 45 jansen , r.a . ,
fabricant , d. , franx , m. , & caldwell , n. 2000 , , 126 , 331 kennicutt , r.c . & kent , s.m .
1983 , , 88 , 1094 kennicutt , r.c .
1989 , , 344 , 685 kennicutt , r.c . ,
keel , w.c . , & blaha , c.a .
1989 , , 97 , 1022 kennicutt , r.c .
1992 , , 388 , 310 kennicutt , r.c .
1998 , , 36 , 189 kim , d .- c . & sanders , d.b .
1998 , , 119 , 41 kobulnicky , h.a . &
skillman , e.d .
1997 , , 489 , 636 kobulnicky , h.a . ,
kennicutt , r.c . , & pizagno , j.l .
1999 , , 514 , 544 krger , h. , fritze - v .
alvensleben , u. , & loose , h .- h .
1995 , , 303 , 41 larkin , j.e .
, armus , l. , knop , r.a . , soifer , b.t .
, & matthews , k. 1998 , , 114 , 59 leitherer , c. & heckman , t.m .
1995 , , 96 , 9 leitherer , c. , schaerer , d. , goldader , j.d . , delgado , r.m.g . ,
robert , c. , kune , d.f . ,
de mello , d.f . ,
devost , d. , & heckman , t.m .
1999 , , 123 , 3 lequeux , j. , peimbert , m. , rayo , j.f . ,
serrano , a. , & torres - peimbert , s. 1979 , , 80 , 155 lequeux , j. , le bourlot , j. , des forets , g.p . , roueff , e. , boulanger , f. , & rubio , m. 1994 , , 292 , 371 li , a. & draine , b.t .
2001 , , 554 , 778 lu , n. , helou , g. , werner , m.w . ,
dinerstein , h.l .
, dale , d.a . ,
silbermann , n.a . ,
malhotra , s. , beichman , c.a . , &
jarrett , t. 2003 , , 588 , 199 malhotra , s. , kaufman , m. , hollenbach , d. et al .
2001 , , 561 , 766 maloney , p.r .
, hollenbach , d.j . , &
tielens , a.g.g.m .
1996 , , 466 , 561 martin , c.l . ,
kobulnicky , h.a . , & heckman , t.m .
2002 , , 574 , 663 massey , p. & armandroff , t.e .
1995 , , 109 , 2470 mcgaugh , s.s .
1991 , , 380 , 140 mcgaugh , s.s .
1994 , , 426 , 135 moorwood , a.f.m .
& oliva , e. 1988 , , 203 , 278 murphy , t.w . , matthews , k. , & soifer , b.t . 1999 , , 763 , 1176 oke , j.b . & gunn , j.e .
1983 , , 266 , 713 pilyugin , l.s . &
ferrini , f. 1998 , , 336 , 103 pradhan , a.k . &
zhang , h.l .
1993 , , 409 , l77 quinet , p. , le dourneuf , m. , & zeippen , c.j .
1996 , , 120 , 361 reunanen , j. , kotilainen , j.k . ,
laine , s. , & ryder , s.d .
2000 , , 529 , 853 rho , j. , reynolds , s.p . ,
reach , w.t . , jarrett , t.h . ,
allen , g.e . , & wilson , j.c .
2003 , , 592 , 299 roche , p.f . ,
aitken , d.k . ,
smith , c.h . , &
james , s.d .
1986 , , 218 , 19 rodrguez , m. & rubin , r.h .
2003 , proceedings of iau symposium 217 , sydney , australia , july 2003 , in preparation roussel , h. , sauvage , m. , vigroux , l. , bosma , a. , bonoli , c. , gallais , p. , hawarden , t. , madden , s. , & mazzei , p. 2001
, , 372 , 406 roussel , h. , helou , g. , beck , r. , condon , j.j . , matthews , k.y . , & jarrett , t.h .
2003 , , in press rubin , r.h . ,
simpson , j.p . ,
haas , m.r . , & erickson , e.f .
1991 , , 103 , 834 rubin , r.h .
1997 , , 474 , l131 sankrit , r. et al .
1998 , , 504 , 344 schlegel , d.j . ,
finkbeiner , p.f . , &
davis , m. 1998 , , 500 , 525 schlegel , e.m . ,
blair , w.p . , & fesen , r.a .
2000 , , 120 , 791 smith , d.a . ,
herter , t. , haynes , m.p . , & neff , s.g .
1999 , , 510 , 669 smutko , m.f . &
larkin , j.e .
1999 , , 117 , 2448 spoon , h.w.w . , keane , j.v . , tielens , a.g.g.m . ,
lutz , d. , & moorwood , a.f.m .
2001 , , 365 , 353 stone , r.p.s . 1977 , , 218 , 767 storey , p.j . &
hummer , d.g .
1995 , , 272 , 41 taylor , c.l . ,
httemeister , s. , klein , u. , & greve , a. 1999 , , 349 , 424 terlevich , r. , tenorio - tagle , g. , france , j. , & melnick , j. 1992 , , 255 , 713 thronson , h.a . , hunter , d.a . ,
casey , s. , & harper , d.a .
1990 , , 355 , 94 vacca , w.d . , garmany , c.d . , & shull , j.m . 1996 , , 460 , 914 vanzi , l. , hunt , l.k . ,
thuan , t.x . , & izotov , y.i .
2000 , , 363 , 493 volker , s. 2000 , , 312 , 859 waller , w.h .
1991 , , 370 , 144 wilcots , e.m . &
miller , b.w .
1998 , , 116 , 2363 lcccccccc c [ tab : sample ] ic 10 & ibm ? & 002024.5 & @xmath35591730 & 0.8 & @xmath31348 & @xmath310.48 & 8.1 & 8.17 + ngc 0693 & i0 : sp & 015031.0 & @xmath35060842 & 22.1 & 1567 & @xmath350.13 & 9.8 & 8.4 + ugc 2855 & sb(s)cd ii - iii & 034822.6 & @xmath35700757 & 18.7 & 1202 & @xmath350.31 & 10.4 & 8.4 + ngc 1569 & ibm ; sbrst & 043049.0 & @xmath35645053 & 2.5 & @xmath31104 & @xmath310.66 & 8.7 & 8.19 + ngc 2388 & sa(s)b : pec & 072853.5 & @xmath35334905 & 54.8 & 4134 & @xmath351.11 & 10.9 & + ngc 4418 & ( r@xmath84)sab(s)a & 122654.6 & @xmath31005240 & 27.3 & 2179 & @xmath351.12 & 10.6 & + ngc 6946 & sab(rs)cd & 203452.3 & @xmath35600914 & 4.5 & 48 & @xmath310.34 & 9.8 & 9.28 + ngc 7218 & sb(r)c & 221011.7 & @xmath31163936 & 23.8 & 1662 & @xmath310.17 & 9.7 & 8.6 + ngc 7771 & sb(s)a & 235124.8 & @xmath35200642 & 60.0 & 4277 & @xmath350.41 & 11.1 & 9.1 + llcccc [ tab : obs ] c 2000 sep 16/17 & * ngc 6946 * & 203452.3 @xmath35600914 & , , , & 09 & photometric + & * ngc 1569 nw*&043047.3 @xmath35645101 & & 09 & photometric + 2000 sep 17/18 & * ngc 6946 ne*&203506.0 @xmath35601057 & , , , & 06 & thin cirrus + & ngc 1569 nw & 043047.3 @xmath35645101 & & 06 & photometric + 2000 sep 18/19 & * ngc 6946
n * & 203452.3 @xmath35600918 & , , , & 07 & thin cirrus + 2000 oct 13/14 & * ngc 6946 s * & 203452.3 @xmath35600910 & , & 08 & thin cirrus + & ngc 1569 nw & 043047.3 @xmath35645101 & , , & 08 & thin cirrus + 2000 oct 14/15 & ngc 6946 s & 203452.3 @xmath35600910 & , , & 07 & photometric + & ngc 6946 & 203452.3 @xmath35600914 & & 07 & photometric + & * ngc 1569 se*&043051.7 @xmath35645048 & & 06 & photometric + 2001 jan 06/07 & ngc 1569 se & 043051.7 @xmath35645048 & & 05 & thin cirrus + & * ngc 2388 * & 072853.5 @xmath35334908 & , , , & 05 & thin cirrus + & * ngc 4418 * & 122654.6 @xmath31005240 & , , , & 06 & thin
cirrus + 2001 oct 03/04 & * ngc 7218 * & 221011.7 @xmath31163939 & , , , & 08 & photometric + & * ugc 2855 * & 034820.4 @xmath35700758 & , , & 09 & photometric + 2001 oct 04/05 & * ngc 7771 * & 235124.8 @xmath35200642 & , , , & 09 & photometric + & * ic 10 * & 002027.2 @xmath35591743 & , , & 12 & photometric + 2001 oct 05/06 & ic 10 & 002027.2 @xmath35591743 & & 16c & thin cirrus + & * ngc 0693 * & 015030.9 @xmath35060842 & , , , & 07 & thin cirrus + 2001 nov 05/06 & ugc 2855 & 034820.4 @xmath35700758 & & 08 & photometric + lcccccccc c [ tab : optical_spectroscopy ] ngc 0693&sep 1996 & 40 90&@xmath3113.6&[email protected]&[email protected]&@[email protected]&@[email protected]&@[email protected] + ugc 2855&sep 1997 & 50&@xmath3114.9&2.0 @xmath56.2 & & @[email protected] & & + ngc 1569&sep 1997 & 20&@xmath3112.9&1.1 @xmath56.1 & & @[email protected] & & + ngc 4418&mar 1996 & 40 45 & & & & & & + ngc 7218&sep 1996 & 30 90&@xmath3114.3&0.5 @xmath56.1 & [email protected]&@[email protected]&@[email protected]&@[email protected] + ngc 7771&sep 1996 & 60 60&@xmath3113.8&1.1 @xmath56.1 & [email protected]&@[email protected]&@[email protected]&@[email protected] + lcrrrrrr [ tab : fluxes ] c ic 10 & & 1400&&200 & @xmath851.5 & @xmath850.7 & 17 + & & & & 190 7 & @xmath855.0 & @xmath852.5 & 59 2 + ngc 0693 & nucleus & 96&&3.9 & 4.5 & 3.7 & 5.4 + & & & & 1.10.1 & 1.80.1 & 2.70.2 & 3.70.2 + ugc 2855 & nucleus & 34 & & 90 .
& 41 & 8.9 & 31 + & & & & 32 2 & 19 1 & 3.60.2 & 24 2 + ngc 1569 nw&&1500&6.6 & 420 & 9.1 & 6.4 & 100 + & & & 1.80.1&100 2 & 3.60.2 & 5.20.3 & 82 3 + ngc 1569 se & & 400 & & 55 & & & 14 + & & & & 30.80.7&&&21.90.4 + ngc 2388 & nucleus & 180&18 & 68 & 26 & 6.7 & 12 + & & & 3.50.1&18.50.7&13.90.8&2.80.6 & 5.30.8 + ngc 4418 & nucleus & 5.5&&@xmath850.5 & 0.7 & 16 & @xmath851.1 + & & & & @xmath850.2 & 0.50.1 & 13.90.5&@xmath850.9 + ngc 6946 & nucleus & 290&88 & 180 & 75 & 61 & 64 + & & & 2.80.1&9.00.2 & 10.50.3&6.50.2 & 7.00.2 + ngc 6946 ne & & 50&@xmath851.6 & 22 & & @xmath852.1 & 8.5 + & & & @xmath8513 & 41 4 & & @xmath8523 & 14040 + ngc 7218 & nucleus & 66&&8.0 & 1.9 & @xmath850.5 & 1.2 + & & & & 4.80.4 & 1.80.2 & @xmath850.7 & 2.40.7 + ngc 7771 & nucleus & 160 & & 40 & 12 & 26 & 15 + & & & & 7.30.3 & 9.70.8 & 16 2 & 5.80.4 + @xmath86c & & 0.7834&0.2903&0.2815&0.1907&0.1247&0.1204 + lccccccc [ tab : results ] c ic 10 & & @xmath87&2.39 , 0.02.7 & 50.5 & 1.77 & 0.6 & @xmath851.4 + ngc 0693 & nucleus & @xmath88&0.17,11.74.4 & 53.1 & 4.35 & 0.8 & 25 + ugc 2855 & nucleus & @xmath89&2.45 , 3.70.8 & 53.4 & 4.72 & 0.7 & 26 + ngc 1569 nw&&@xmath90&2.32 , 0.30.9 & 51.9 & 3.23 & 0.4 & 0.82 + ngc 1569 se&&@xmath91&2.32 , 0.03.8 & 51.1 & 2.41 & 0.5 & + ngc 2388 & nucleus & @xmath92&0.19 , 2.30.8 & 53.8 & 5.05 & 0.7 & 19 + ngc 4418 & nucleus & @xmath93&0.08 , & @xmath8552.7&@xmath853.98&@xmath940.9&@xmath9423 + ngc 6946 & nucleus & @xmath95&1.13 , 3.10.9 & 52.3 & 3.62 & 0.7 & 14 + ngc 6946 ne&&@xmath96&1.13 , 2.80.9 & 51.4 & 2.69 & 0.5 & @xmath851.1 + ngc 7218 & nucleus & @xmath88&0.11 , 0.60.9 & 51.9 & 3.19 & 0.8 & 20 . + ngc 7771 & nucleus & @xmath97&0.24 , 3.00.9 & 53.8 & 5.12 & 0.7 & 12 + | the palomar integral field spectrograph was used to probe a variety of environments in nine nearby galaxies that span a range of morphological types , luminosities , metallicities , and infrared - to - blue ratios . for the first time
, near - infrared spectroscopy was obtained for nuclear or bright regions in star - forming galaxies over two spatial dimensions ( 57 @xmath0 100 ) in the ( 1.257 ) , ( 1.644 ) , ( 1.282 ) , ( 2.122 ) , and ( 2.166 ) transition lines .
these data yield constraints on various characteristics of the star - forming episodes in these regions , including their strength , maturity , spatial variability , and extinction .
the regions stand out from the nuclei .
unlike observations of nuclear regions , region near - infrared observations do not show a spatial coincidence of the line and continuum emission ; the continuum and line maps of regions usually show distinct and sometimes spatially - separated morphologies . gauging from and equivalent widths and luminosities , the regions have younger episodes of star formation than the nuclei and more intense radiation fields .
near - infrared line ratio diagnostics suggest that regions have `` purer '' starbursting properties .
the correlation between ionizing photon density and mid - infrared color is consistent with the star formation activity level being higher for regions than for nuclei . and
though the interpretation is complicated , on a purely empirical basis the regions show lower fe@xmath1 abundances than nuclei by an order of magnitude . |
a celebrated invariant of a knot @xmath3 in 3-space is its _
function _ @xmath4 defined for complex numbers of absolute value @xmath5 , and taking values in the set of integers .
the signature function of a knot is a concordance invariant , and plays a key role in the study of knots via surgery theory , @xcite .
it turns out that the signature function is a locally constant function away from the ( possibly empty ) set @xmath6 of roots of the alexander polynomial on the unit circle . in view of this , the interesting part of the signature function is its _ jumping behavior _ on the set @xmath7 . in other words
, we may consider the associated _ jump function _
@xmath8 defined by @xmath9 .
we may identify the jump function with a _ jump divisor _
@xmath10 $ ] in @xmath11 .
since @xmath12 and @xmath13 , it follows that the jump function uniquely determines the signature function away from the set @xmath7 . since @xmath14 , it follows in particular that @xmath15 determines the _ _ of the knot @xmath16 .
the signature of a knot may be defined using a seifert surface of a knot ( see section [ sub.symmetries ] below ) .
an intrinsic definition of the jump function of a knot was given by milnor @xcite , using the blanchfield pairing of the universal abelian cover of a knot .
this definition , among other things , makes evident the role played by the roots of the alexander polynomial on the unit circle ( as opposed to the rest of the roots of the alexander polynomial , which are ignored ) . from the point of view of gauge theory and mathematical physics ,
the signature function of a knot may be identified with the spectral flow of a 1-parameter family of the signature operator , twisted along abelian ( that is , @xmath0-valued ) representations of the knot complement . the moduli space of @xmath0 representations of the knot complement is well understood ; it may be identified with the unit circle .
on the other hand , the moduli space of @xmath1 representations is less understood , and carries nontrivial topological information about the knot and its dehn fillings , as was originally discovered by casson ( see @xcite ) and also by x - s .
lin ; see @xcite .
one may ask to identify the @xmath0 representations which _ deform to _ irreducible @xmath1 representations . using a linearization argument ,
klassen and frohman showed that a necessary condition for a @xmath0 representation @xmath17 to deform is that @xmath18 .
this brings us to the ( square of the ) set @xmath7 .
conversely , frohman - klassen proved sufficiency provided that the alexander polynomial has _ simple roots _ on the unit circle ; see @xcite .
herald proved sufficiency under the ( more relaxed condition that ) the jump function vanishes nowhere ; see @xcite .
it is unknown at present whether sufficiency holds without any further assumptions .
let us summarize the two key properties of the jump divisor @xmath19 , in the spirit of mazur ( see @xcite ) : * the jump divisor controls the signature function of a knot .
* the jump divisor controls ( infinitesimally ) deformations of @xmath0 representations of the knot complement to irreducible @xmath1 representations .
it is a long standing problem to find a formula for the function of a knot in terms of its _ colored jones function_. the latter is a sequence of jones polynomials associated to a knot .
recall that given a knot @xmath3 and a positive integer @xmath20 ( which corresponds to an @xmath20-dimensional irreducible representation of @xmath21 ) , one can define a laurrent polynomial @xmath22 $ ] . in @xcite , rozansky considered a repackaging of the sequence @xmath23 .
namely , he defined a sequence of rational functions @xmath24 for @xmath25 with the following properties : * @xmath26 for some polynomials @xmath27 $ ] with @xmath28 and such that @xmath29 .
* for every @xmath20 we have : @xmath30\ ] ] where @xmath31 $ ] is the ring of formal power series in @xmath32 with rational coefficients equation is often called the _ euler expansion _ of the colored jones function . in physical terms , the above expansion is an asymptotic expansion of the chern - simons path integral of the knot complement , expanded around a backround @xmath0 flat connection .
thus , philosophically , it should not be a surprise to discover that this expansion has something to do with the signature of the knot .
for the curious reader , let us point out that rozansky conjectured such an expansion for the full kontsevich integral of a knot , graded by the negative euler characteristic of graphs ( thus the name , euler expansion ) .
this conjecture was proven by kricker and the author ; @xcite .
furthermore , a close relation was discovered between residues of the rational functions @xmath33 at roots of unity and the lmo invariant of cyclic branched coverings of the knot ; @xcite . in an attempt to understand the euler expansion , a theory of finite type invariants of knots ( different from the usual theory of vassiliev invariants )
was proposed in @xcite . according to that theory , two knots are @xmath34-equivalent iff they are @xmath35-equivalent ; @xcite .
moreover , @xmath33 is a finite type invariant of type @xmath36 .
technically , the euler expansion of the colored jones function is an integrality statement .
namely , it is easy to see that there exist unique sequence of power series @xmath37 $ ] for @xmath25 that satisfies equation .
the hard part is to show that these power series are taylor series expansions of rational functions with integer coefficients and prescribed denominators .
the statement @xmath28 in the leading term of the euler expansion is nothing but the melvin - morton - rozansky conjecture , proven by bar - natan and the author in @xcite .
thus , the leading order term in the euler expansion is a well - understood topological invariant of knots .
ever since the euler expansion was established , it has been a question to establish a topological understanding of the lower order terms .
consider @xmath38 where @xmath39 .
we will think of @xmath40 as a function ( with singularities ) defined on the unit circle .
if @xmath41 is a root of the alexander polynomial on @xmath11 , we may expand @xmath42 around @xmath43 .
the result is a power series with lowest term @xmath44 , for some integer @xmath45 and some nonzero real number @xmath46 .
let us define the _ jones jump function _ of a knot @xmath3 @xmath47 by @xmath48 where @xmath49 is the imaginary part of a complex number @xmath50 and @xmath51 is the _ sign _ of a real number @xmath52 is defined by @xmath53 or @xmath54 according to @xmath55 or @xmath56 respectively .
we say that a knot @xmath3 is _ simple _ if its alexander polynomial @xmath57 has simple roots on the unit circle .
if @xmath3 is simple , then @xmath58 .
a modest corollary is : if @xmath3 is simple , conjecture [ conj.1 ] implies that the colored jones function of @xmath3 determines the signature @xmath16 .
notice that @xmath59 and @xmath60 .
thus , it suffices to check the conjecture on the upper semicircle .
the conjecture is false if @xmath57 has multiple roots ( of odd or even multiplicity ) .
for example , consider the connected sum @xmath61 of @xmath20 right trefoils .
then , @xmath62 and @xmath63 .
we present the following evidence for the conjecture : @xmath64 conjecture [ conj.1 ] is true for torus knots , and for knots with at most 8 crossings .
+ @xmath65 the conjecture is compatible with the operations of mirror image , connected sum ( assuming the resulting knot is simple ) and @xmath66 parallels of knots . en route to establish our results , we give a skein formula that uniquely characterizes the jump function of simple knots ; see theorem [ thm.unique ] .
let us compare conjecture [ conj.1 ] with existing conjectures about the structure of the colored jones function .
at the time of the writing , there are two conjectures that relate the colored jones function to hyperbolic geometry .
namely , * the _ hyperbolic volume conjecture _ , after kashaev and j&j.murakami , which states that for a hyperbolic knot @xmath3 , @xmath67 where @xmath68 . * the _ characteristic equals deformation variety conjecture _ , due to the author , which compares the deformation curve of @xmath69 representations of a knot complement ( viewed from the boundary ) with a complex curve which is defined using the recursion relations ( with respect to @xmath20 ) of the sequence @xmath23 ; see @xcite and @xcite .
the hyperbolic volume conjecture is an analytic statement , which involves the existence and identification of a sequence of real numbers . on the other hand ,
the characteristic equals deformation variety conjecture is an algebraic statement , since it is equivalent to the equality of two polynomials with integer coefficients , one of which is obtained by noncommutative elimination , and the other obtained by commutative elimination .
conjecture [ conj.1 ] appears to be an analytic conjecture , since its basic ingredients are signs of real numbers . in the field of quantum topology
, analytic conjectures have held the longest .
let us end the introduction with the following understand the underlying geometry and perturbative quantum field theory behind the taylor expansion of the @xmath70 function ( and more generally , euler expansion of the colored jones function ) .
in particular , use the higher order terms @xmath33 in the expansion to formulate a conjecture for the jump function of all knots .
the author wishes to thank s. orevkov , l. rozansky and a. stoimenov , and especially j. levine for helpful conversations .
given a seifert matrix @xmath71 of a knot @xmath3 , consider the _
hermitian matrix _
@xmath72 , for @xmath73 .
the eigenvalues of @xmath74 are real , and we define @xmath75 , where @xmath76 denotes the _ signature _ of a hermitian matrix @xmath77 .
it turns out that @xmath78 is independent of the seifert surface @xmath71 chosen . since @xmath79 , where @xmath80 , and @xmath81 is the _
symmetrized alexander polynomial _ of @xmath3 , it follows that @xmath78 is a locally constant function with possible jumps along the set @xmath7 . the next lemma , which follows from the proof of ( * ? ? ?
* corollary 2 ) , summarizes the symmetries of the jump function . if @xmath17 is a root of the alexander polynomial on @xmath11 , then @xmath82 , where * @xmath83 is an integer * @xmath84 , where @xmath85 is the multiplicity of @xmath17 in @xmath57 , and * @xmath86 .
moreover , @xmath87 .
in particular , if @xmath3 is simple , @xmath15 takes values in the set @xmath88 . for a precise formula for the jump function in that case , see theorem [ thm.jump ] .
let us begin with a useful definition .
a triple of links @xmath89 is called _ bordered _ if there is an embedded ball @xmath90 in @xmath91 that locally intersects them as in figure [ crossing ] .
@xmath92 if we choose planar projection and a crossing , then a bordered triple corresponds to replacing the crossing by a positive , negative or smoothening . notice that if @xmath93 is a link with @xmath5 component , then @xmath94 and @xmath95 are links with @xmath5 and @xmath96 components respectively .
the next lemma computes the change of the signature function with respect to the change of a crossing , in terms of the sign of the alexander polynomials . if @xmath3 is a knot , @xmath97 such that @xmath98 , then @xmath99 we can choose seifert surfaces @xmath100 for @xmath101 such that @xmath102 where @xmath103 and @xmath104 are some row vectors .
hermitianizing , we get : @xmath105 let us call a triple of hermitian matrices @xmath106 _ @xmath17-bordered _ if @xmath107 for @xmath108 and some row vector @xmath109 .
using lemma [ lem.tbordered ] the result follows .
if @xmath106 is a @xmath17-bordered triple , and @xmath110 , then @xmath111 this is well - known for @xmath112 ; @xcite and also ( * ? ? ?
* lemma 3.1 ) .
we give a proof here for all @xmath17 . by a similarity transformation ( that is a replacement of @xmath113 by @xmath114 where @xmath115 is an invertible matrix , and @xmath116 is the conjugate transpose of @xmath115 ) , we can assume that @xmath117^r \oplus d , \ ] ] where @xmath118 is a nonsingular diagonal matrix , @xmath119^r$ ] is the zero @xmath120 matrix , @xmath109 is a @xmath121 vector and @xmath122 a real number . since the nullity ( that is , the dimension of the kernel ) and the signature of the matrix @xmath123 are given by : [ cols="<,^,^,^",options="header " , ] the result follows by a case - by - case argument .
the next theorem computes the jump function of a simple knot in terms of a relative sign of alexander polynomials .
first , a preliminary definition .
if @xmath124 is a real - valued analytic function of @xmath52 in a neighborhood of @xmath122 , we define _ the sign of @xmath125 at @xmath122 _ @xmath126 to be the sign of the first nonvanishing taylor series coefficient ( around @xmath122 ) , if there is such , and zero otherwise . in other words ,
if @xmath127 , we have : @xmath128 where @xmath129 for @xmath130 and @xmath131 .
notice that if @xmath132 , then @xmath133 , and that if @xmath122 is a simple root , then @xmath134 where @xmath135 is sufficiently small and positive .
fix a simple knot @xmath3 and a complex number @xmath136 .
choose a planar projection of @xmath3 and a crossing ( positive or negative ) .
then , @xmath137 , where @xmath138 is the sign of the chosen crossing .
suppose that @xmath139 .
such a projection and choice of crossing will be called @xmath140-_good_. fix @xmath140 as above . for every @xmath140-good projection
, we have @xmath141 without loss of generality , let us assume @xmath142 , that is @xmath143 .
we will apply lemma [ lem.skeins ] twice to @xmath144 and @xmath145 for sufficiently small positive @xmath135 . under these assumptions
, we have that @xmath146 ( since @xmath17 is an isolated root of a polynomial ) and @xmath147 ( since @xmath148 by assumption ) , and similarly for @xmath149 .
thus , the hypothesis of lemma [ lem.skeins ] are satisfied . applying lemma [ lem.skeins ] twice , we get @xmath150 and @xmath151 now , subtract and remember that @xmath152 is continuous at @xmath17 since @xmath153 .
we get @xmath154 since @xmath3 is simple , it follows that @xmath155 , thus the cases @xmath156 or @xmath157 do not occur above .
thus , @xmath158 the result follows using remark [ rem.signf ] .
indeed , @xmath159 and @xmath160 .
there is a unique invariant @xmath161 defined for a simple knot @xmath3 and @xmath162 such that for every @xmath140-good projection we have : @xmath163 in view of theorem [ thm.jump ] , we need to prove that there is at most one such invariant . fix a simple knot @xmath3 and a complex number @xmath136 .
we need to prove that there exists a @xmath140-good projection .
start with any planar projection of @xmath3 and a crossing .
if it is not good , apply reidemaster moves ii , which frohman - klassen call _ threading _ and improve it to be good , using the proof of ( * ? ? ?
* theorem 6.2 ) .
thus , conjecture [ conj.1 ] is equivalent to the following : @xmath64 for every simple knot @xmath3 , and every @xmath164 , we have @xmath165 .
+ @xmath65 moreover , for every @xmath140-good projection we have : @xmath166
in this section we will prove conjecture [ conj.1 ] for torus knots .
let @xmath167 denote the @xmath168 _ torus knot _
, where @xmath169 are coprime natural numbers .
for example , @xmath170 is the right - hand trefoil .
the alexander polynomial of torus knots is given by : @xmath171 the roots of @xmath172 on the unit circle are @xmath173 complex roots of unity which are not @xmath122 or @xmath174 order roots of unity .
they are all simple . using a useful parametrization of them , following kearton ( * ? ? ?
* sec.13 ) , we obtain that latexmath:[\[\mathrm{roots}_{\d(t_{a , b})}=\{t(m , n):=e^{2 \pi i ( m / a+n / b ) } \ , @xmath59 , we need only compute the jump at the points @xmath176 where @xmath177 , @xmath178 and @xmath179 . in @xcite kearton
computes the jump function of torus knots by @xmath180 in other words , we have : @xmath181 now we discuss the @xmath70 function of torus knots , which was originally computed by rozansky ( see ( * ? ? ?
* eqn.(2.2 ) ) ) , and most recently , it has been recomputed by march and ohtsuki ; see @xcite .
we understand that bar - natan has unpublished computations of the euler expansion of the kontsevich integral of torus knots . according to (
* eqn.(2.2 ) ) , the @xmath70 function of torus knots is given by : @xmath182 given an analytic function @xmath183 let us define @xmath184 we have that @xmath185 when we expand @xmath186 around a root @xmath41 , only the last two terms of the numerator contribute to the coefficient of @xmath187 .
that is , @xmath188 now , suppose that @xmath183 is a laurrent polynomial with real coefficients that satisfies @xmath189 .
then , @xmath190 .
thus , @xmath191 and if we substitute @xmath192 , we get @xmath193 if @xmath194 is a simple root of @xmath195 on the unit circle ( as is the case for the alexander polynomial of torus knots ) , then the above real number is negative .
this proves that @xmath196 and confirms conjecture [ conj.1 ] for torus knots .
let @xmath125 denote either the @xmath70 function or the function of a knot .
the following list describes some well - known properties of @xmath125 . *
if @xmath197 denote the knot @xmath3 with opposite orientation , then @xmath198 .
* if @xmath199 denote the mirror image of @xmath3 , then @xmath200 .
* if @xmath201 denotes the connected sum of knots , then @xmath202 . *
if @xmath203 denote the @xmath66 parallel of a knot @xmath3 with zero framing , then @xmath204 .
the stated behavior of the signature function under @xmath66 parallel was proven by kearton @xcite , and for the @xmath70 function was proven by ohtsuki ( * ? ? ?
3.1 ) . from this
, it follows that if conjecture [ conj.1 ] is true for a simple knot @xmath3 , then it is true for @xmath197 , @xmath199 , @xmath203 ( for all @xmath20 ) .
furthermore , if @xmath205 is simple , and conjecture [ conj.1 ] is true for @xmath206 and @xmath207 , then it is also true for @xmath205 . in this section
we will verify conjecture [ conj.1 ] by computer calculations .
rozansky has written a maple program that computes the @xmath70 function of a knot ; see @xcite .
we will use a minor modification qfunction.mws of rozansky s program , adopted for our needs . in qfunction.mws
, the knot is described by a braid word .
for example , @xmath208 $ ] represents the braid @xmath209 whose closure is the @xmath210 knot in classical notation .
the command @xmath211)$ ] gives a list whose first , second and third entries are the braid word , the polynomials @xmath212 and @xmath57 , where @xmath213 .
a sample output of the program is : .... > # the right trefoil 3_1 > br1([1,1,1 ] ) ; > 2 2 4 [ [ 1 , 1 , 1 ] , 1 + z , 2 z + z ] > # the 4_1 knot > br1([1,-2,1,-2 ] ) ; > 2 [ [ 1 , -2 , 1 , -2 ] , 1 - z , 0 ] > # the 7_2 knot > br1([-1,3,3,3,2,1,1,-3,2 ] ) ; > 2 2 4 [ [ -1 , 3 , 3 , 3 , 2 , 1 , 1 , -3 , 2 ] , 1 + 3 z , 12 z + 14 z ] > # 7_3 > br1([1,1,2,-1,2,2,2,2 ] ) ; > 2 4 [ [ 1 , 1 , 2 , -1 , 2 , 2 , 2 , 2 ] , 1 + 5 z + 2 z , 2 4 6 8 22 z + 65 z + 46 z + 9 z ] .... for example , for the right hand trefoil , we have : @xmath214 the mathematica program jjump.m computes the @xmath215 function .
for example , we may launch the jjump.m program from a mathematica session . ....
( math100)/home / stavros : math mathematica 5.0 for sun solaris ( ultrasparc ) copyright 1988 - 2003 wolfram research , inc . --
motif graphics initialized -- in[1]:= < < jjump.m in[2]:= poles[1+z^2,2z^2+z^4 ] solve::ifun : inverse functions are being used by solve , so some solutions may not be found ; use reduce for complete solution information .
out[2]= { { 0.16666666666666666667 , -0.00844343197019481429 } } .... we learn that the coefficient of @xmath216 of @xmath217 ( where @xmath218 is the right trefoil ) around the root @xmath219 , is @xmath220 .
this computes that @xmath221 , as needed . similarly , ....
in[4]:= poles[1 + 5z^2 + 2z^4,22z^2 + 65z^4 + 46z^6 + 9z^8 ] solve::ifun : inverse functions are being used by solve , so some solutions may not be found ; use reduce for complete solution information .
out[4]= { { 0.075216475230034463796 , -0.00388836700144941422 } , > { 0.27241752919082620707 , -0.00542424178920663095 } } .... we learn that the coefficient of @xmath222 of @xmath223 around the roots @xmath224 and @xmath225 are @xmath226 and @xmath227 respectively .
this computes the jump function @xmath228 for @xmath229 .
now , let us compute the jump function of a knot
. in @xcite orevkov gives a mathematica program sm.mat which takes as input a braid presentation of a knot , and gives as output a seifert surface of a knot .
launching the jump.m version of it in a mathematica session produces ....
( math100)/home / stavros : math mathematica 5.0 for sun solaris ( ultrasparc ) copyright 1988 - 2003 wolfram research , inc . --
motif graphics initialized -- in[1]:= < < jump.m in[2]:= jump[{1,1,1 } ] inversefunction::ifun : inverse functions are being used .
values may be lost for multivalued inverses .
solve::ifun : inverse functions are being used by solve , so some solutions may not be found ; use reduce for complete solution information .
out[2]= { -2 } .... which computes the jump function on the upper semicircle for the right trefoil @xmath218 . ....
in[3]:= jump[{1,1,2,-1,2,2,2,2 } ] inversefunction::ifun : inverse functions are being used .
values may be lost for multivalued inverses .
solve::ifun : inverse functions are being used by solve , so some solutions may not be found ; use reduce for complete solution information .
out[3]= { -2 , -2 } .... which computes the jump function on the upper semicircle for the @xmath230 knot .
this confirms the conjecture for the @xmath218 and @xmath230 knots . in the appendix
, we give the source code of two mathematica programs , jump.m and jjump.m which compute the @xmath231 and the @xmath215 function of knots .
.... ( * poles[p , ap ] computes the poles of the rational functions p / ap^2 * ) ( * at the roots of ap=0 on the unit circle .
p , ap are polynomials in z * ) ( * poles2[p , ap ] lists the coefficients of the taylor expansion at * ) ( * ( t - a)^{-2}. * ) ( * poles[p , ap ] lists { a , coefficient of taylor expansion at ( t - a)^{-2 } } * ) ff[x_]:=x[[2 ] ] ; poles[ap_,p_]:=module [ { quotient , apt , roots , poles , k } , quotient = simplify[p / ap^2 /.
( z->z^{1/2 } ) /.
( z->2 cos[2*pi*t]-2 ) ] ; apt= simplify[ap /.
( z->z^{1/2 } ) /.
( z->2 cos[2*pi*t]-2 ) ] ; roots = select[map[ff , flatten [ nsolve[apt = = 0 , t , 20 ] ] ] , 1/2 > # > 0 & ] ; poles= { } ; table[flatten[{roots[[k ] ] , coefficient[series[quotient,{t , roots[[k]],0 } ] , t - roots[[k]],-2 ] } ] , { k , length[roots ] } ] ] ( * for the 3_1 knot : poles[1+z^2,2z^2+z^4 ] * ) ( * for the 4_1 knot : poles[1-z^2,0 ] * ) ( * for the 7_2 knot : poles[1 + 3z^2,12z^2 + 14z^4 ] * ) ( * for the 7_3 knot : poles[1 + 5z^2 + 2z^4,22z^2 + 65z^4 + 46z^6 + 9z^8 ] * ) ....
.... ( * computing the signature and jump function of knots presented as * ) ( * closures of braids . * ) ( * the signature of the right trefoil is signaturebraid[{1,1,1}]=-2 * ) ( * signaturem[a ] of a matrix a is the signature of a+a^ * * ) ( * jump[{1,1,1 } ] is the jumps of the signature of the right trefoil * ) jump[brd_]:=module [ { m , v , aps , hermitian , roots , k } , m = max[abs @ brd]+1 ; v = n[seifertmatrix[m , brd ] ] ; hermitian=(1-exp[2*pi*i*s])v+(1-exp[-2*pi*i*s ] ) transpose[v ] ; aps = n[det[(cos[2*pi*s/2]+i sin[2*pi*s/2])v-(cos[2*pi*s/2]-i sin[2*pi*s/2 ] ) transpose[v]],20 ] ; roots = select[map[ff , flatten [ nsolve[{aps = = 0 , im[s]==0 } , s , 15 ] ] ] , 1/2 > # > 0 & ] ; if[length[roots]==0 , { } , flatten[table[signaturem[hermitian /.
s->(roots[[k]]+1/1000 ) ] -signaturem[hermitian /.
s->(roots[[k]]-1/1000 ) ] , { k , length[roots ] } ] ] ] ] ( * 7_3 knot signaturebraid[{1,1,2,-1,2,2,2,2 } ] * ) ( * 7_5 knot signaturebraid[{1,1,1,1,2,-1,2,2 } ] * ) ( * 8_2 knot signaturebraid[{-1,2,2,2,2,2,-1,2 } ] * ) ( * 8_5 knot signaturebraid[{1,1,1,-2,1,1,1,-2 } ] * ) ( * 8_15 knot signaturebraid[{1,1,-2,1,3,3,2,2,3 } ] * ) ( * 7_3 , 7_5 , 8_2 , 8_5 , 8_15 have signature -4 * ) .... | the signature function of a knot is a locally constant integer valued function with domain the unit circle . the jumps ( i.e. , the discontinuities ) of the signature function can occur only at the roots of the alexander polynomial on the unit circle .
the latter are important in deforming @xmath0 representations of knot groups to irreducible @xmath1 representations . under the assumption that these roots are simple , we formulate a conjecture that explicitly computes the jumps of the signature function in terms of the jones polynomial of a knot and its parallels . as evidence , we prove our conjecture for torus knots , and also ( using computer calculations ) for knots with at most @xmath2 crossings .
we also give a formula for the jump function at simple roots in terms of relative signs of alexander polynomials . |
there is still no agreement about the nature of the low - temperature phase of the ising spin glass , which is defined by the hamiltonian @xmath5 the spins can take the values @xmath6 , and the nearest - neighbour couplings @xmath7 are independent from each other and are most often chosen to be gaussian distributed with mean zero and a standard deviation @xmath8 . while many monte - carlo simulations show properties conforming to the replica - symmetry - breaking ( rsb ) scenario ( implying many low - temperature states and a lack of self - averaging ) @xcite , other simulations @xcite and analytical arguments @xcite favour the droplet picture ( a scaling theory based on the existence of only one low - temperature state and its time reverse ) .
the ambiguities stem from the difficulty in reaching the asymptotic limit of low temperatures and large system sizes .
monte - carlo results are likely to be affected by finite - size and critical - point effects .
we have recently shown that a system that is known to conform to the droplet picture at sufficiently large system sizes has features similar to those of rsb if only small systems are studied and if the temperature is not low enough @xcite .
this system is the hierarchical lattice , or , equivalently , the migdal - kadanoff approximation ( mka ) applied to a cubic or hypercubic lattice .
it is thus possible that the ising spin glass on three- or four - dimensional lattices might show its true low - temperature properties only beyond the length scales accessible to present - day monte - carlo simulations .
exact evaluation of ground states and low - lying excited states appears to indicate a scenario that agrees neither with the droplet picture nor with the rsb theory , but shows instead low - lying excitations which are fractal @xcite .
newman and stein have argued @xcite that such excitations can not exist in the thermodynamic limit .
as the studied system sizes are very small , the phenomenon might be a small - size effect that vanishes at larger system sizes .
since fractal excitations are not possible on a hierarchical lattice ( only combinations of compact droplets and domain walls can occur on it ) , the mka can not show these low - lying excitations , and agrees with the droplet picture even for small system sizes at low temperatures with a gaussian distribution for the bonds @xmath7 .
very recently several papers have focussed on the @xmath0 ising spin glass , where the nearest - neighbour couplings take only the values 1 and @xmath9 , instead of being chosen from a gaussian distribution .
evidence is accumulating that the ground - state degeneracy introduces new effects .
thus , krzakala and martin @xcite argued that even if a system showed rsb at low temperatures , different valleys in the energy landscape would differ in entropy to the extend that for sufficiently large system sizes one state would dominate the zero - temperature partition function , leading for instance to a trivial overlap distribution ( i.e. an overlap distribution that is the sum of two @xmath10-functions at opposite values of the overlap ) .
this argument is supported by simulations by palassini and young @xcite who find a crossover from a zero - temperature behaviour with a trivial overlap distribution to a finite - temperature behaviour which seems to agree with the rsb scenario .
in contrast , he d , hartmann and domany , claim to find a non - trivial overlap distribution even at zero temperature @xcite .
it is the purpose of this paper to study the low temperature properties of the @xmath0 model in mka in order to shed some light on the results of monte - carlo simulations , and to determine the conditions under which the true low - temperature behaviour should be visible .
our findings confirm the conjecture by krzakala and martin that the zero - temperature behaviour is different from the low - temperature behaviour , and they also confirm the scaling assumptions concerning the entropy differences used in their argument . furthermore ,
our results show that the true asymptotic zero - temperature behaviour and the true low - temperature behaviour can be seen only beyond the length scales currently studied with monte - carlo simulations .
the outline of this paper is as follows : in section [ simu ] we present our numerical results for the overlap distribution , the binder parameter , and the recursion of the couplings within mka . in section
[ theory ] , we give scaling arguments that yield the asymptotic exponents and the crossover behaviour seen in the simulations .
section [ concl ] summarizes and discusses the results .
the migdal - kadanoff approximation is a real - space renormalization group the gives approximate recursion relations for the various coupling constants .
evaluating a thermodynamic quantity in mka in @xmath11 dimensions is equivalent to evaluating it on an hierarchical lattice that is constructed iteratively by replacing each bond by @xmath12 bonds , as indicated in fig .
the total number of bonds after @xmath13 iterations is @xmath14 .
@xmath15 , the smallest non - trivial system that can be studied , corresponds to a system linear dimension @xmath16 , @xmath17 corresponds to @xmath18 , @xmath19 corresponds to @xmath20 and so on .
note that the number of bonds on hierarchical lattice after @xmath13 iterations is the same as the number of sites of a @xmath11-dimensional lattice of size @xmath21 .
thermodynamic quantities are then evaluated iteratively by tracing over the spins on the highest level of the hierarchy , until the lowest level is reached and the trace over the remaining two spins is calculated @xcite .
this procedure generates new effective couplings , which have to be included in the recursion relations .
the recursion relation of the width @xmath22 of the two - spin coupling is for sufficiently many iterations and sufficiently low temperature given by @xmath23 , with @xmath24 in mka in three dimensions ( which is the only dimension studied in this paper ) .
= 0.2 we first evaluated the overlap distribution @xmath25 , \label{p}\ ] ] between two identical replicas of the system , where the superscripts @xmath26 and @xmath27 denote the two replicas of the system , @xmath28 is the number of bonds of a system of size @xmath29 , and @xmath30 and @xmath31 $ ] denote the thermodynamic and disorder average respectively . as discussed in @xcite , the calculation of @xmath32
is made easier by first calculating its fourier transform @xmath33 , which is given by @xmath34 .
\label{fy}\ ] ] the recursion relations for @xmath33 involve two- and four - spin terms , and can easily be evaluated numerically because all terms are now in an exponential .
having calculated @xmath33 , one can then invert the fourier transform to get @xmath32 .
figure [ fig2 ] shows our results for @xmath35 at different temperatures , and for @xmath36 at different system sizes respectively . due to the ground state degeneracy , the overlap distribution for fixed
@xmath29 does not change below a temperature for which most samples are in the ground state . with increasing system size ,
the peaks in the overlap distribution become sharper , and the probability of finding an overlap value near zero decreases , indicating that one state and its spin - flipped counterpart dominate the statistics .
= figure [ fig3 ] shows the probability density @xmath37 that the two replicas have zero overlap , at several different temperatures .
the two curves for @xmath36 and @xmath38 are on top of each other , indicating that at these sizes and temperatures , the system is in the ground state with a probability close to 1 .
the @xmath39 curve coincides for @xmath40 with the @xmath1 curve , but branches off for larger @xmath29 and approaches the slope -0.26 expected from the droplet picture and seen in a system with gaussian distributed couplings ( see @xcite ) .
the @xmath41 curve seems to be affected by ground state effects for @xmath42 , as it starts out close to the @xmath1 curve and then has a negative slope which becomes flatter for larger @xmath29 .
a slope flatter than that predicted by the droplet picture indicates an influence of the critical point , as discussed in @xcite . for even larger @xmath29
, the curve must become steeper again and approach the slope of the droplet picture .
even the @xmath43 curve appears to be affected by the ground state degeneracy for @xmath44 . from these numerical results , the asymptotic behaviour of the @xmath1 curve can not be predicted .
it seems unlikely that it becomes flatter for larger @xmath29 , implying that it is fundamentally different from the droplet picture , which should govern the behaviour of a sufficiently large system at low temperatures .
the asymptotic slope of the @xmath1 curve and the crossover length scale at which a finite - temperature curve branches off from it will be derived further below .
furthermore , our results show that at system sizes smaller than around 16 , the asymptotic low - temperature behaviour is not visible for any temperature , as the influence of the critical point and of the ground state are too strong .
= 0.6 in order to be able to study larger system sizes , we determined the binder parameter @xmath45}{3[\langle q^2 \rangle]^2 } \right ) , \label{binder}\ ] ] which can be obtained by differentiating eq .
( [ fy ] ) with respect to @xmath46 .
this is done by evaluating @xmath33 for 3 small values of @xmath46 .
the systematic error resulting from the finiteness of @xmath46 is found by evaluating @xmath47 for a few samples for many values of @xmath46 , and by extrapolating to @xmath48 .
the binder parameter @xmath49 in the high - temperature phase , and approaches 1 in the low - temperature phase if the overlap distribution is trivial . within the droplet picture ,
@xmath50 must scale as @xmath51 for sufficiently large @xmath29 .
figure [ fig4 ] shows our results . as for the overlap distribution ,
the @xmath1 curve is much steeper than the limit slope expected from the droplet picture , and the low - temperature curves branch off from it at a system size that is larger for lower temperatures .
in contrast to figure [ fig3 ] , system sizes can be studied that are large enough to see the differences between the three curves for @xmath36 , 0.2 and 0 .
the @xmath1 simulation was done by taking the trace only over those configurations that contribute to the ground state , and by keeping track of the degeneracies .
= 0.6 next , we tried to understand the reasons for the steep decline of the @xmath1 curve , and of its slow approach to the asymptotic slope . for sufficiently large system sizes ,
the main contribution to @xmath37 at @xmath1 must come from samples where a domain wall costs no energy .
a domain wall is introduced into the system by flipping one of the two corner spins of the hierarchical lattice out of the ground state orientation , and by determining the new ground state resulting with this boundary condition .
samples where such a domain wall costs no energy have zero effective coupling strength @xmath52 at length scale @xmath29 , with @xmath22 resulting from the recursion relation for the width of the distribution of the couplings under the renormalization procedure . figure [ fig5 ] shows the probability of having @xmath52 ( or , equivalently , of having a domain wall with zero energy cost ) as function of the system size @xmath29 .
one can see that the slope is identical to that expected from the droplet picture beyond length scales @xmath53 , and is only slightly steeper for smaller system sizes .
this indicates that the @xmath0 model has a crossover length around 32 , which is not present in the model with gaussian distributed couplings , where the slope agrees with the droplet picture even for the smallest system sizes .
since figure [ fig5 ] agrees with the droplet picture , it can not explain the steep decrease of @xmath37 , and @xmath54 at zero temperature .
for small system sizes , not only domain walls , but also several small droplets can create a zero overlap , but this effect should become irrelevant for sufficiently large system sizes .
the only remaining possibility is that while the probability for having a domain wall of zero energy agrees with the droplet picture , the weights of the two ground states with and without a domain wall differ by a factor that increases with increasing system size . in order to check this hypothesis
, we evaluated the degeneracies of the two ground states that are obtained by fixing the two corner spins in parallel and antiparallel orientation respectively , and derived from this the probability that two identical replicas of the system are in the two different states .
the result is shown in figure [ fig6 ] .
the curve is steeper by @xmath9 compared to that of the droplet picture , indicating that entropy differences between ground states are the crucial factor causing the deviation of the @xmath1 results from the droplet picture .
= 0.6 = 0.6 figure [ fig6 ] shows the same slow approach towards asymptopia as figures [ fig3 ] and [ fig4 ] . we will attempt an explanation in the next section which is devoted to a theoretical explanation of the numerical findings .
the main objective of the scaling theory presented here is to derive the asymptotic slope of the @xmath1 curves , and to predict the crossover length scale at which curves at finite @xmath55 branch off from the @xmath1 curve .
as shown by our numerical data presented in the previous section , entropy differences between ground state configurations that differ by a domain wall play a crucial role .
let us therefore consider a system that has a ground state for which a domain wall costs no energy , and let us estimate the order of magnitude of the entropy difference between the two ground states .
one of the ground states is obtained by fixing the two corner spins of the system ( those with the highest coordination number ) in parallel orientation , and the other is obtained by making them antiparallel .
contributions to the entropy of each of these two states are made by droplet excitations that cost no energy .
( a droplet is a block of spins that are connected to each other and that does not include one of the two corner spins , and it may comprise just a single spin ) . by flipping several droplets
, one can thus get from every configuration contributing to one of the two ground states ( with fixed corner spins ) to every other configuration contributing to this state .
the argument made in the following is similar to the one made by krzakala and martin @xcite for a system with supposed rsb in the low - temperature phase . to each configuration contributing to the first ground state ( with parallel corner spins ) ,
there exist configurations in the second ground state ( with antiparallel corner spins ) that differ from it only by a domain wall .
this means that the two configurations can be transformed into each other by flipping a coherent block of spins including the right corner spin ( assuming that the left corner spin is up in both states ) .
now , all the possible zero - energy droplets in the first configuration that do not touch the domain wall , are also zero - energy droplets in the second configuration . droplets that do not touch domain walls can therefore make no contribution to the entropy difference between the two ground states , because they occur in both of them .
the entropy difference between the two states results therefore from those droplets that touch the domain wall .
now , the domain wall involves @xmath56 bonds , where @xmath4 is the fractal dimension of the domain wall , and has the value @xmath57 in mka .
the average number of droplets touching the domain wall can therefore be expected to be @xmath56 ( assuming that the majority of droplets are small and independent from each other ) , and the typical fluctuation ( measured over different samples , or over the two ground states ) in the number of droplets touching a domain wall can be expected to be @xmath58 , which is identical to @xmath29 in mka in three dimensions .
now , the probability that two replicas have zero overlap is proportional to the probability that a domain wall costs no energy , @xmath59 , multiplied by the probability that the configurations of the two replicas have a different relative orientation of the corner spins , @xmath60 .
this explains the asymptotic slope of @xmath61 seen in figure [ fig6 ] , and expected for the zero - temperature @xmath37 curve in figure [ fig3 ] . for the binder parameter , figure [ fig4 ] , we expect the same asymptotic behaviour , @xmath62 .
the reason is that for large system sizes mainly samples with zero - energy domain wall excitations show a considerable difference between @xmath63 and @xmath64 .
next , let us discuss possible reasons why the asymptotic slope @xmath65 is approached so slowly in all our plots .
our scaling argument is based on the assumption that the size distribution of the droplets that touch the domain wall does not change much with the system size . for small system sizes
, the droplet size distribution might be far from the asymptotic droplet size distribution , possibly causing considerable deviations from asymptopia .
this effect is probably more severe in mka than on a three - dimensional lattice , because in mka droplets consisting of a single spin with coordination number two make no contribution to the entropy difference between the two ground states .
the reason is that flipping the domain wall transforms each spin next to the domain wall with coordination number 2 that can be flipped without energy cost ( and which therefore makes a contribution to the entropy ) into a spin that can be flipped only by paying the energy 4 , while every spin with coordination number two along the domain wall that can be flipped only by paying energy , is transformed into a spin that can be flipped without energy cost .
the numbers of the two classes of spins must therefore be equal , and the entropy contribution due to spins with coordination number two that can be flipped without energy cost is the same for both ground states .
furthermore , we have assumed that the fluctuation in the numbers of droplets touching the domain wall is given by the central limit theorem , which is a good approximation only for sufficiently large system sizes .
deviations from the numbers predicted by the central limit theorem may be a further reason why the asymptotic slope @xmath65 is only visible for large system sizes .
third , we have assumed that domain walls make the main contribution to @xmath37 .
this assumption is not correct for small system sizes , where droplet excitations that do not involve the corner spin may also add up to an overlap value of zero .
finally , let us determine the crossover length scale beyond which the droplet picture should become visible for small nonzero temperatures : within the droplet picture , we have @xmath66 , while we have at zero temperature @xmath67 .
a crossover between the two regimes occurs when the two quantities are equal , i.e. , when @xmath68 .
in this paper , we have studied the @xmath0 ising spin glass within mka .
we have found that the zero - temperature behaviour is fundamentally different from that at low temperatures , due to entropy differences between ground states . only for length scales larger than of the order @xmath3
does the expected droplet - picture behaviour become visible .
we have presented a scaling theory that predicts the asymptotic scaling exponent @xmath65 for the overlap distribution at zero temperature , and we have shown from numerical results as well as from analytical arguments that the approach to this asymptotic scaling might be slow .
our findings shed some light on recent monte - carlo simulations of the three - dimensional @xmath0 ising spin glass . while the scaling arguments by krzakala and martin @xcite predict a decrease of @xmath37 at zero temperature at least with an exponent @xmath69 ( if one assumes with them that the system shows rsb , implying @xmath70 ) , which lies somewhere between -1.1 and -1.3 , the best monte - carlo simulations find only a value around -0.9 @xcite .
other monte - carlo simulations giving a considerably smaller exponent probably do not sample the ground state configurations with the appropriate weights ( see the comment by marinari et al @xcite on the simulations by hatano and gubernatis @xcite , and the remarks by palassini and young @xcite on the simulations by hartmann @xcite . ) our findings of a surprisingly slow approach to the correct asymptotic scaling can reconcile the monte - carlo results with the predictions by krzakala and martin , and also with our predictions based on the droplet picture ( where the asymptotic exponent is around -1.4 or -1.5 ) , which we believe to be the correct description of the spin - glass phase .
our results in figure [ fig3 ] show also that for not too low temperatures the overlap distribution data may at first ( for the smallest @xmath29 values ) be affected by the ground - state degeneracy ( as indicated by a slope that is initally steeper than for larger @xmath29 ) , then ( for somewhat larger @xmath29 ) by the critical point ( manifesting itself in a pretty flat slope ) , and only for sufficiently large @xmath29 ( which may be beyond the reach of monte - carlo simulations ) the correct asymptotic slope given by the droplet picture . given such a complicated behaviour , the predictions in @xcite and @xcite for the asymptotic behaviour of @xmath32 based on small system sizes and assuming simple scaling forms have no convincing basis .
we conclude that it is possible that the @xmath0 ising spin glass in three dimensions and for system sizes smaller than approximately 16 does not show the correct asymptotic scaling behaviour at any value of the temperature .
it remains to be seen whether the ising spin glass with a gaussian bond distribution has also finite - size effects which make it impossible to see even at low temperatures the correct asymptotic scaling behaviour for the system sizes presently used in computer simulations .
e. marinari , g. parisi , f. ricci - tersenghi , and j.j .
ruiz - lorenzo in ( _ spin glasses and random fields _
, ed : a. p. young ) , ( world scientific , singapore , 1997 ) .
e. marinari , g. parisi , f. ricci - tersenghi , j. ruiz - lorenzo , f. zuliani , j. stat . phys . * 98 * , 973 ( 2000 ) .
m. palassini and a. p.
young , phys
lett . * 83 * , 5126 ( 1999 ) .
newman and d.l.stein , phys .
rev . lett . * 76 * , 515 ( 1996 ) ) ; * 76 * , 4821 ( 1996 ) .
m. a. moore , h. bokil and b.drossel , phys .
. lett . * 81 * , 4252 ( 1998 ) .
h. bokil , b. drossel and m.a .
moore , phys .
e * 62 * , 946 ( 2000 ) .
m. palassini and a.p .
young , phys .
* 85 * , 3017 ( 2000 ) .
f. krzakala and o.c .
martin , phys .
* 85 * , 3013 ( 2000 ) .
newman and d.l .
stein , cond - mat/0010033 .
f. krzakala and o.c .
martin , europhys . lett . * ? ?
m. palassini and a.p .
young , cond - mat/0012161 .
g. he d , a.k .
hartmann and e. domany , cond - mat/0012451 .
southern and a.p .
young , j. phys .
c * 10 * , 2179 ( 1977 ) .
berg , u.e .
hansmann and t. celik , phys .
b * 50 * , 16 444 ( 1994 ) .
e. marinari , g. parisi , f. ricci - tersenghi and f. zuliani , cond - mat/0011039 .
n. hatano and j.e .
gubernatis , cond - mat/0008115 .
a.k . hartmann , eur .
j. b * 13 * , 539 ( 2000 ) . | we study the low - temperature phase of the three - dimensional @xmath0 ising spin glass in migdal - kadanoff approximation . at zero temperature , @xmath1
, the properties of the spin glass result from the ground - state degeneracy and can be elucidated using scaling arguments based on entropy .
the approach to the asymptotic scaling regime is very slow , and the correct exponents are only visible beyond system sizes around 64 . at @xmath2 , a crossover from the zero - temperature behaviour to
the behaviour expected from the droplet picture occurs at length scales proportional to @xmath3 where @xmath4 is the fractal dimension of a domain wall .
canonical droplet behaviour is not visible at any temperature for systems whose linear dimension is smaller than 16 lattice spacings , because the data are either affected by the zero - temperature behaviour or the critical point behaviour . 2 |
neutrino interactions with plasma are very important to understand supernova explosions and many other astrophysical phenomena @xcite
. two types of effects arise from such interactions .
first , they modify the neutrino flavor oscillations @xcite , and introduce a resonant coupling between different flavor states , known as the msw ( mikheyev - smirnov - wolfenstein ) effect @xcite .
second , they create an induced neutrino charge @xcite , which can lead to collective plasma oscillations and significantly increase the collision cross sections . the energy transfer between a neutrino beam and plasma wave
is mediated by the neutrino landau damping @xcite .
the individual flavor processes are important to understand the solar neutrino deficit , while the collective plasma effects could play a major role in supernova explosions @xcite .
it should be noticed that the core - collapse problem is still unsolved @xcite . a possible solution could eventually be given by these plasma effects . in a recent work ,
we have proposed to built a bridge between these two kinds of phenomena , and have introduced plasma physics methods in the discussion of neutrino flavor oscillations in matter . in particular , we have derived exact bgk ( bernstein - greene - kruskal ) like solutions for electron plasma density profiles compatible with given neutrino flavor parameters @xcite , and have determined the modified neutrino flavor oscillations in the presence of plasma waves and turbulence @xcite . here , we take a further step in the same direction , by considering the mutual influence of flavor oscillations and plasma instabilities . we will show that the neutrino flavor parameters are not only perturbed by electron plasmas oscillations , but can also contribute to both the dispersion relation and the growth rates of neutrino beam instabilities in a plasma .
this paper is organized in the following way . in section
ii , we consider the basic equations of our problem , by considering a simple fluid description where ions are assumed at rest and the neutrino flavor oscillations are taken into account . in section iii , we consider a plasma in steady state , and characterize the unperturbed solutions for both the electron plasma parameters and the flavor polarization vector . in section iv , we consider the perturbations induced by a plasma wave and establish the evolution equations for the perturbed quantities . from this perturbative analysis
it becomes clear that plasma waves induce perturbation in the neutrino flavor parameters which , on the other hand , lead to new dispersive effects . in section
v , we consider the case of electron plasma waves excited by an incoherent neutrino beam , and determine the corresponding growth rates . in section
vi we generalize this analysis to the case of a coherent neutrino beam and show that the flavor oscillations can contribute to the dispersion relation of electron plasma waves .
finally , in section vii , we state some conclusions .
we assume a simple fluid description for both the plasma electrons and the neutrino beam , with immobile ions . a unit system with @xmath0 will be used .
the electrons are described by the non - relativistic fluid equations fluid equations @xmath1 where @xmath2 and @xmath3 are the electron mean density and velocity , @xmath4 and @xmath5 is the electron pressure .
we have also used the lorentz force @xmath6 , and the neutrino force @xmath7 , defined by @xmath8 where @xmath9 and @xmath10 are the electric and magnetic fields , @xmath11 is the electron charge , @xmath12 the fermi constant of weak interactions , and @xmath13 and @xmath14 are effective fields induced by the weak interactions , determined by @xmath15 the electron - neutrino density @xmath16 , and current @xmath17 , are coupled to the muon neutrino density @xmath18 and current @xmath19 , as shown by the continuity equations @xmath20 where the quantity @xmath21 pertains to neutrino coherence , as clarified below . here ,
for simplicity , we restrict our analysis to the familiar two - flavor model @xcite , but extension to the three neutrino flavor states would not be difficult .
in these two coupled equations we have also used the constants @xmath22 strictly speaking , eq .
( [ 2.4 ] ) only apply to a coherent neutrino beam with a defined energy @xmath23 , and can not be used in arbitrary situations . for this reason
, we will only restrict our discussion to the neutrino beam interaction with a plasma , where the frequency @xmath24 can be unequivocally defined as @xmath25 , and @xmath26 is the neutrino square mass difference .
on the other hand , the quantum coherence factor @xmath21 satisfies the relations @xmath27 where we have defined @xmath28 the meaning of the total time derivative in eq .
( [ 2.6 ] ) will be clarified in section iii . to complete the description of the neutrino populations
, we should consider the neutrino momentum equations @xmath29 with @xmath30 and @xmath31 .
the quantities @xmath32 and @xmath33 appearing in these equations are effective fields @xcite defined as @xmath34 the quantity @xmath35 in eq .
( [ 2.5 ] ) depends on the energy of the neutrino beam , and is well defined for a ( nearly mono - energetic ) beam , such that @xmath36 .
the momentum equations in ( [ 2.8 ] ) show that , in the presence of plasma perturbations @xmath37 and @xmath38 , the quantities @xmath39 and @xmath40 are not necessarily identical . in conclusion , we have a rather detailed model for the neutrino - plasma coupling .
the electron variables @xmath2 and @xmath3 are determined in a self - consistent way together with the fields @xmath41 through the maxwell equations with self - consistent charge and current density , with a coupling produced by the neutrino force @xmath7 in the electron momentum equation ( [ 2.1 ] ) .
the neutrino force depends on the effective neutrino fields @xmath42 , which in turn are specified by the electron neutrino variables @xmath43 .
however , in the present two - flavor model , the electron neutrino variables oscillate due to the quantum coherence @xmath21 , coupling with the muon neutrino quantities @xmath44 .
finally , the neutrino oscillations are influenced by the plasma in two ways : due to the coupling with @xmath2 in @xmath45 in eq .
( [ 2.7 ] ) and due to the role of the electromagnetic field in the electron neutrino momentum equation ( [ 2.8 ] ) . in the following ,
illustrative examples of applications are provided .
for simplicity , let us consider a non - magnetized plasma , although e.g. regarding a supernovae setting future works accounting for a non - zero ambient magnetic field could be quite relevant . in equilibrium , we have @xmath46 , @xmath47 and @xmath48 .
this implies that , for a mono - energetic neutrino beam , the velocity of both flavors are identical , and we can use @xmath49 . in this case , eq .
( [ 2.4 ] ) will reduce to the following relation @xmath50 with @xmath51 . in this case
, the coupled equations ( [ 2.6 ] ) and ( [ 3.1 ] ) can describe the evolution of a three - dimensional flavor polarization vector @xmath52 , where the total time derivatives are defined without ambiguity .
they can be rewritten as @xmath53 with @xmath54 .
this implies that @xmath55 introducing a new angle @xmath56 , we can also write @xmath57 equation ( [ 3.3 ] ) can be solved as @xmath58 it is convenient to define the constants of integration as @xmath59 , in terms of a free parameter @xmath60 related to quantum coherence , leading to @xmath61 accordingly , without loss of generality a phase constant was chosen so that @xmath62 . for completeness
, we show the results for the remaining polarization vector components , compatible with the normalization condition @xmath63 , @xmath64 where the angle @xmath65 satisfy @xmath66 for instance , in the case of absence of quantum coherence ( @xmath67 ) one has @xmath68 .
we can explicitly see that @xmath60 is a measure of the amplitude of the flavor oscillations , for if @xmath67 there are no oscillations at all .
the corresponding solution for the equilibrium electron - neutrino density is @xmath69 \ , , \label{3.6}\ ] ] and @xmath70 .
these results are strictly valid for a mono - energetic beam @xmath71
. a finite value of @xmath72 would introduce a temporal ( and/or spatial ) decay of these oscillations @xcite .
nothing that eq .
( [ 3.2 ] ) contain total derivatives ( in a lagrangian variables sense ) , we can also consider another class of initial conditions , @xmath73 and @xmath74 .
the resulting solutions would then be expressed in terms of spatial coordinates as @xmath75 where @xmath76 . in the following ,
we will focus on equilibrium solutions of the type ( [ 3.5 ] ) , although a similar analysis could be done for solutions of the type ( [ 3.7 ] ) .
in the previous section there was no electromagnetic field at all .
we now consider the possible excitation of electron plasma waves by a mono - energetic neutrino beam , as described by the perturbed densities @xmath77 , @xmath78 , and @xmath79 .
we restrict our discussion to electrostatic waves , with @xmath80 and @xmath9 determined by poisson s equation @xmath81 linearizing the electron fluid eqs .
( [ 2.1 ] ) and ( [ 2.2 ] ) , with @xmath82 , we get @xmath83 in addition , the displacement current contribution from ( [ 2.3 ] ) has been disregarded due to the non - relativistic assumption . taking the time derivative of the continuity equation and using ( [ 4.1 ] )
we obtain @xmath84 where we have used the electron plasma frequency @xmath85 , and the electron thermal velocity @xmath86 , where @xmath87 is the electron temperature . in this expression
, we can easily recognize the usual wave equation for electron density perturbations , with an additional term associated with the electron - neutrino oscillations . in order to determine the quantity @xmath88 , we linearize eqs .
( [ 2.4 ] ) - ( [ 2.7 ] ) , assuming that the unperturbed neutrino flavor solutions @xmath89 , @xmath90 and @xmath91 are given by the ( homogeneous in space ) solutions of section iii .
this leads to @xmath92 complemented by the neutrino momentum equations @xmath93 in these expressions we have used the total time derivative @xmath94 .
we can immediately recognize that the muon - neutrinos are not accelerated by the electron plasma wave , because they are not directly coupled to the electron perturbations .
this is strictly valid only for a neutral electron - proton plasma .
we should also note that @xmath95,where @xmath23 is the unperturbed neutrino beam energy . from here , we can derive an evolution equation for the perturbed electron - neutrino density , of the form @xmath96 with @xmath97 it can be seen that the evolution of the perturbed density @xmath88 is coupled to @xmath98 , and also depends on the perturbed coherence @xmath99 . an equation for this quantity
can be derived from eq .
( [ 2.6 ] ) , and can be written as @xmath100 where it was defined @xmath101 . using the above equations ( [ 4.4 ] ) and ( [ 4.4b ] ) , and taking the definition of @xmath102 into account
, we can then transform this into @xmath103 + \frac{\omega_0}{n_0 } n_{e0 } \nabla \cdot \tilde { \bf v}_e \ , , \label{4.7}\ ] ] where the perturbed neutrino velocity is determined by the equation of motion @xmath104 .
we now have all the equations for the perturbed quantities , which will be solved in the next two sections .
let us first consider the simple case of @xmath105 and @xmath67 , which corresponds to the absence of quantum coherence . in this case , we have two coupled equations for the variables @xmath88 and @xmath98 . on the other hand , no flavor oscillations will occur , and @xmath106 .
we therefore take @xmath107 . for perturbations evolving as @xmath108 , we can reduce eqs .
( [ 4.3 ] ) and ( [ 4.5 ] ) to @xmath109 and @xmath110 from here , we can derive the dispersion relation for plasma waves in the presence of an incoherent neutrino beam , as @xmath111 this can also be written in the standard form , by introducing the plasma dielectric function @xmath112 , as @xmath113 where the electron and neutrino susceptibilities are defined by the expressions @xmath114 this dispersion relation shows the possible occurrence of neutrino beam instabilities .
maximum growth rates occur for the double resonance condition @xmath115 , where we have assumed @xmath116 .
the corresponding growth rate for the unstable solution is @xmath117 valid for @xmath118 and high frequency waves such hat @xmath119 . in the growth rate expression ,
the free parameters are @xmath120 ( the initial electron neutrino population ) , @xmath23 ( the defined coherent neutrino beam energy ) and the neutrino beam velocity @xmath121 as well as the equilibrium electron density .
finally , it should be observed that eq .
( [ 5.3 ] ) is valid for @xmath122 : it can easily be shown from eq .
( [ 5.1b ] ) that there is no instability if @xmath123 , as expected .
hence , the above growth rate estimate is always defined for the relevant scenarios . also observe the growth rate is @xmath124 which is the same estimate as in @xcite for a weak neutrino beam .
revealing insights can be obtained rewriting the dispersion relation in terms of a characteristic function @xmath125 as @xmath126 the dispersion relation is a fourth degree equation for @xmath127 .
hence , to admit complex conjugate solutions and hence an unstable mode , the local minimum at @xmath128 ( see figure 1 ) should satisfy @xmath129 . to
leading order in the neutrino coupling effects , we find @xmath130 together with the instability condition @xmath131 where @xmath132 was also assumed .
notice that the instability is enhanced by larger neutrino effects as apparent from the second term in eq .
( [ ff ] ) .
moreover if the double resonance condition is fulfilled the instability condition is also immediately satisfied since the neutrino term is always positive ( never stabilizing ) . ) , in a generic unstable case such that @xmath133 , where @xmath134 is the local minimum .
there are only two real roots for the dispersion relation equation . on the right : the same but for a generic stable case such that @xmath135 and four real roots for the dispersion relation.,width=576,height=249 ]
let us now consider the case of a coherent and nearly mono - energetic neutrino beam , where we have to retain the contributions of the coherence parameter @xmath21 . to enhance this contribution , we suppose @xmath136 in the unperturbed solutions ( [ 3.5b])-([full ] ) .
we can then rewrite the coupled equations ( [ 4.5 ] ) and ( [ 4.7 ] ) in the following form @xmath137 and @xmath138 \right\ } \tilde n = 0\ , , \label{6.2}\ ] ] with the following auxiliary quantities @xmath139 to simplify the analysis , we neglected the terms containing the perturbed electron and muon neutrino velocities .
we then assume perturbations of the form @xmath140 with @xmath141 , this allows us to establish the following relations @xmath142 @xmath143 and @xmath144 from here , we obtain the recurrence relation @xmath145 \ , \label{6.5}\ ] ] where @xmath146 is given by eq .
( [ 5.2 ] ) , with @xmath127 replaced by @xmath147 , and the coupling function is defined by @xmath148 } \label{6.6}\ ] ] in order to study mode coupling contributions , we can use an approximation in eq .
( [ 6.5 ] ) , by noting that the differences between @xmath149 and @xmath150 are very small , if we assume that @xmath151 .
( [ 6.4a ] ) , we obtain by differentiation , @xmath152 .
this leads to @xmath153 now , using @xmath154 , and noting that @xmath155 , we finally arrive at the simple estimate @xmath156 taking @xmath157 , this leads to the following dispersion relation @xmath158 this result shows that the previous dispersion relation ( [ 5.2 ] ) for the neutrino beam interactions is corrected by an additional factor due to quantum correlations , which is of the order of @xmath159 .
such corrections result from the coupling between quantum flavor oscillations and electron plasma waves , or in other words , between the quantum properties of neutrinos and their collective interactions with the plasma .
another interesting case , would be that of a near resonance between plasma and flavor oscillations , such that @xmath160 , where stronger effects can be expected .
we have studied the influence the neutrino flavor oscillations on electron plasma waves created by neutrino beams , or in other words , the influence of the neutrino quantum properties on their collective behavior .
the flavor oscillations result from the difference between the neutrino mass states and their interaction states .
on the other hand , the neutrino beam instabilities result form their weak coupling with the plasma .
we have shown that the existence of flavor oscillations gives a new contribution to plasma dispersion relation , thus changing the frequency and growth rates of the beam instabilities . in our model
, we have introduced a number of simplifying assumptions .
first , we have only retained the charged weak current and ignored the contributions from the neutral weak current .
it is known that the electron neutrinos are coupled by the charged bosons @xmath161 to the electrons , and all neutrino flavors are coupled by the neutral boson @xmath162 with both electrons and protons . for a plasma of electrons and protons in equilibrium , this weak coupling would give no net contribution to the neutrino - plasma interactions .
in contrast , in the presence of a perturbation , this would lead to a correcting factor of order one to the terms proportional to @xmath12 .
furthermore , our work was only focused on electron plasma oscillations , and the protons ( or ions ) were assumed immobile .
but we could also consider the excitation of ion acoustic waves , with frequencies of the order of the flavor oscillation frequency .
these excitation could be driven by flavor oscillations .
we should notice that the collective ion - neutrino coupling is mediated by the electrons , which can be assumed in boltzmann equilibrium in the electrostatic and weak field potentials .
an electronic density @xmath164 m@xmath165 @xcite yields a plasma frequency @xmath166 s@xmath167 .
considering also @xmath168 m@xmath165 , @xmath169 mev , @xmath170 and with @xmath171 j.m@xmath172 , one finds a growth rate , @xmath173 which is far shorter than the time scale ( @xmath174 second ) of the explosion .
this instability definitely appears fast enough to alter neutrino mixing in core - collapse supernova .
the present results show that electron plasma waves excited by intense neutrino beams are intimately linked with the quantum processes associated with flavor oscillations .
the dispersion relation and growth rates of these plasma instabilities are directly influenced by these flavor oscillations .
it should however be noticed that we have also restricted our analysis to nearly mono - energetic neutrino beams .
the general case of an arbitrary neutrino population can only be treated in the frame of a quantum statistical approach , which will be considered in a future publication . | we consider the collective neutrino plasma interactions , and study the electron plasma instabilities produced by a nearly mono - energetic neutrino beam in a plasma .
we describe the mutual influence of neutrino flavor oscillations and electron plasma waves .
we show that the neutrino flavor oscillations are not only perturbed by electron plasmas waves , but also contribute to the dispersion relation and the growth rates of neutrino beam instabilities . |
it is by now well established that the physics associated with classical and quantum fields in curved spacetimes can be reproduced , within certain approximations , in a variety of different physical systems the so - called `` analogue models of general relativity ( gr ) '' @xcite .
the simplest example of such a system is provided by acoustic disturbances propagating in a barotropic , irrotational and viscosity - free fluid . in the context of analogue models
it is natural to separate the kinematical aspects of gr from the dynamical ones . in general , within a sufficiently complex analogue model one can reproduce any pre - specified spacetime and the kinematics of fields evolving on it independently of whether or not it satisfies the classical ( or semiclassical ) einstein equations @xcite .
indeed , to date there are no analogue models whose effective geometry is determined by einstein equations . in this sense
we currently have both analogue spacetimes and analogues of quantum field theory in curved spacetimes , but ( strictly speaking ) no analogue model for gr itself @xcite . in order to reproduce a specific spacetime geometry within an analogue model
, one would have to take advantage of the specific equations describing the latter ( for example , for fluid models , the euler and continuity equations , together with an equation of state ) , plus the possibility of manipulating the system by applying appropriate external forces . in the analysis of this paper
we will think of the spacetime configuration as `` externally given '' , assuming that it has been set up as desired by external means any back - reaction on the geometry is neglected as in principle we can counter - balance its effects using the external forces . in the context of analogue models
this is not merely a hypothesis introduced solely for theoretical simplicity , but rather a realistic situation that is in principle quite achievable .
specifically , in this paper we analyze in simple terms the issue of quantum quasi - particle creation by several externally specified @xmath1-dimensional analogue geometries simulating the formation of black hole - like configurations .
( in a previous companion paper @xcite we investigated the causal structure of these , and other , spacetimes . ) in this analysis we have in mind , on the one hand , the possibility of setting up laboratory experiments exhibiting hawking - like radiation @xcite and , on the other hand , the acquisition of new insights into the physics of black hole evaporation in semiclassical gravity .
all the discussion holds for a scalar field obeying the dalembert wave equation in a curved spacetime .
this means that we are not ( for current purposes ) considering the deviations from the phononic dispersion relations that show up at high energies owing to the atomic structure underlying any condensed matter system .
we shall briefly comment on these modifications at the end of the paper . for simplicity , throughout the paper
we adopt a terminology based on acoustics in moving fluids ( we will use terms such as acoustic spacetimes , sonic points , fluid velocity , etc . ) , but our results are far more general and apply to many other analogue gravity models not based on acoustics .
we summarise the main conclusions below .
first of all , we recover the standard hawking result when considering fluid flows that generate a supersonic regime at finite time .
( that is , we recover a stationary creation of quasi - particles with a planckian spectrum . )
we then analyze the quasi - particle creation associated with other types of configurations .
in particular , we shall discuss in detail a `` critical black hole '' a flow configuration that presents an acoustic horizon without an associated supersonic region . from this analysis
we want to highlight two key results : * the existence of a supersonic regime ( sound velocity @xmath2 strictly smaller than fluid velocity @xmath3 ) is not needed in order to reproduce hawking s stationary particle creation .
we demonstrate this fact by calculating the quantity of quasi - particle production in an evolving geometry which generates only an isolated sonic point ( @xmath0 ) , but without a supersonic region , in a finite amount of laboratory time .
* moreover , in order to produce a hawking - like effect it is not even necessary to generate a sonic point at finite time .
all one needs is that a sonic point develops in the asymptotic future ( that is , for @xmath4 ) _ with sufficient rapidity _ ( we shall explain in due course what we exactly mean by this ) . from the point of view of the reproducibility of a hawking - like effect in a laboratory , the latter result is particularly interesting . in general , the formation of a supersonic regime in a fluid flow normally considered to be the crucial requirement to produce hawking emission is associated with various different types of instability ( landau instability in superfluids , quantized vortex formation in bose
einstein condensates , etc . ) that could mask the hawking effect . to reproduce a hawking - like effect without invoking a supersonic regime
could alleviate this situation . from the point of view of gr
, we believe that our result could also have some relevance , as it suggests a possible alternative scenario for the formation and semiclassical evaporation of black hole - like objects .
the plan of the paper is the following : in the next section we introduce the various acoustic spacetimes on which we focus our attention , spacetimes that describe the formation of acoustic black holes of different types . in section [ sec :
creation ] we present separately the specific calculations of redshift for sound rays that pass asymptotically close to the event horizon of these black holes . by invoking standard techniques of quantum field theory in curved spacetime , one can then immediately say when particle production with a planckian spectrum takes place . finally , in the last section of the paper we summarise and discuss the results obtained .
associated with the flow of a barotropic , viscosity - free fluid along an infinitely long thin pipe , with density and velocity fields constant on any cross section orthogonal to the pipe , there is a ( 1 + 1)-dimensional _ acoustic spacetime _
@xmath5 , where the manifold @xmath6 is diffeomorphic to @xmath7 . using the laboratory time @xmath8 and physical distance @xmath9 along the pipe
as coordinates on @xmath6 , the _ acoustic metric _ on @xmath6 can be written as@xmath10 = \omega^2\left[-c^2\,d t^2 + ( \d x+v\,\d t)^2\right ] \;,\label{metric}\]]where @xmath2 is the speed of sound , @xmath3 is the fluid velocity , and @xmath11 is an unspecified non - vanishing function @xcite . in general , all these quantities depend on the laboratory coordinates @xmath12 and @xmath13 . here
, we shall assume that @xmath2 is a constant .
hence , it is the velocity @xmath14 that contains all the relevant information about the causal structure of the acoustic spacetime @xmath5 .
we direct the reader to the companion paper @xcite for a detailed analysis of the causal structure associated with a broad class of @xmath1-dimensional acoustic geometries , both static and dynamic . the sonic points , where @xmath15 , correspond to the so - called acoustic apparent horizons apparent horizons for the lorentzian geometry defined on @xmath6 by the metric ( [ metric ] ) .
the fact of having an underlying minkowski structure associated with the laboratory observer makes the definition of apparent horizons in acoustic models less troublesome than in gr ( see reference @xcite , pp .
consider a monotonically non - decreasing function @xmath16 such that @xmath17 and @xmath18 for @xmath19 .
if one chooses @xmath20 in ( [ metric ] ) , the corresponding acoustic spacetime represents , for observers with @xmath21 , a static black hole with the horizon located at @xmath22 ( in this case apparent and event horizon coincide ) , a black hole region for @xmath23 , and a ( right - sided ) surface gravity@xmath24we can , moreover , distinguish three cases : * @xmath25 and @xmath26 for @xmath23 : a non - extremal black hole ; * @xmath25 and @xmath27 for @xmath23 : a `` critical '' black hole ; * @xmath28 and @xmath27 for @xmath23 : an extremal black hole .
now , taking the above @xmath16 , let us consider @xmath13-dependent velocity functions@xmath29with @xmath30 a monotonically decreasing function of @xmath13 , such that @xmath31 and @xmath32 .
( the first condition serves to guarantee that spacetime is flat at early times , whereas we impose the second one only for simplicity . all the analysis in the paper
could be performed without adopting this assumption , leaving the physical results unchanged . however , that would require more case - by - case splitting , only to cover new situations without physical interest . )
there are basically two possibilities for @xmath30 , according to whether the value @xmath33 is attained for a finite laboratory time @xmath34 or asymptotically for an infinite future value of laboratory time . in the first case @xmath35 and the corresponding metric ( [ metric ] ) represents the formation of a non - extremal , critical , or extremal black hole , respectively . for small values of @xmath36
we have@xmath37 ^ 2)\;,\label{appxi - bh}\]]where @xmath38 is a positive parameter .
hence the function @xmath30 behaves , qualitatively , as shown in figure [ f : xi - bh ] .
apart from this feature , the detailed behaviour of @xmath30 is largely irrelevant for our purposes . if instead @xmath39 is attained only at infinite future time , that is @xmath40 , one is describing the asymptotic formation of either a critical black hole ( if @xmath41 ; obviously , in this case choosing the non - extremal or the critical @xmath42 profile is irrelevant ) or an extremal black hole ( if @xmath28 ) .
now the function @xmath30 behaves , qualitatively , as shown in figure [ f : xi - cbh ] .
the relevant feature of @xmath43 is its asymptotic behaviour as @xmath44 . in the following
we shall consider two possibilities for this asymptotics , although others can , of course , be envisaged : exponential : @xmath45 , with @xmath46 a positive constant , in general different from @xmath47 , and @xmath48 ; power law : @xmath49 , with @xmath50 and @xmath51 .
for all the situations considered so far , spacetime is minkowskian in the two asymptotic regions corresponding to @xmath52 , and to @xmath4 , @xmath19 ( @xmath53 and @xmath54 , respectively , adopting the notation of reference @xcite ) . starting with a quantum scalar field in its natural minkowskian vacuum at @xmath52 ,
we want to know the total quantity of quasi - particle production to be detected at the right asymptotic region at late times , @xmath4 , caused by the dynamical evolution of the velocity profile @xmath14 . in the geometric acoustic approximation ,
a right - going sound ray is an integral curve of the differential equation@xmath55we are interested in sound rays propagating from @xmath56 ( see figure [ f : conf - bh ] ) ; that is , in solutions of ( [ diffeq ] ) that satisfy an initial condition @xmath57 , with @xmath58 in the limit @xmath59 ( so @xmath60 can be thought of as an `` initial '' event corresponding to the emission of the acoustic signal ) .
if such a ray ends up on @xmath61 , we can identify `` final '' events @xmath62 on it , with @xmath63 as @xmath64 . for a ray connecting @xmath53 to @xmath54 one can also find an event @xmath65 such that @xmath66 , which corresponds to the crossing of the `` kink '' in @xmath3 , located at @xmath67 according to equation ( [ velocity ] ) , by the sound signal . finally , we can define , for such a ray , two parameters @xmath68 and @xmath69 as follows:@xmath70@xmath71such parameters correspond to null coordinates in spacetime .
if an acoustic event horizon @xmath72 is present in the spacetime , the coordinate @xmath68 is regular on it ( , @xmath68 attains some finite value on @xmath72 ) , whereas @xmath69 tends to @xmath73 as @xmath74 is approached .
we can express both @xmath68 and @xmath69 in terms of the velocity profile ( shape and dynamics ) and of the crossing time @xmath75 .
to this end , we can integrate equation ( [ diffeq ] ) , first between @xmath76 and @xmath77:@xmath78then between @xmath77 and @xmath79:@xmath80on replacing @xmath81 from equation ( [ xi - x ] ) into ( [ u ] ) , we find the value of @xmath68 for a generic right - moving ray that crosses the kink at laboratory time @xmath75:@xmath82similarly , substituting @xmath83 from equation ( [ t - t ] ) into ( [ u ] ) , then adding and subtracting the quantity @xmath84 , we find@xmath85 in the analysis below , our chief goal consists of finding the relation between @xmath68 and @xmath69 for a sound ray that is close to the horizon , , in the asymptotic regime @xmath86 .
from such a relation it is then a standard procedure to find the bogoliubov @xmath87 coefficients and hence the total quasi - particle content to be measured , in this case , by an asymptotic observer at @xmath54 .
( see , for example , reference @xcite ) . in the case of an exponential relation between @xmath68 and @xmath69 it is a well established result that a planckian spectrum is observed at late times @xcite , so hawking - like radiation will be recovered .
when the apparent horizon forms at a finite laboratory time , say at @xmath88 , an event horizon always exists , generated by the right - moving ray that eventually remains frozen on the apparent horizon , at @xmath22 .
for such a ray @xmath89 , and since @xmath35 , the @xmath68 parameter has the finite value@xmath90for a ray with @xmath91 we then obtain , combining equations ( [ uu ] ) and ( [ uh - fin]):@xmath92this _ exact _ equation is now in a form suitable for conveniently extracting _ approximate _ results in the region @xmath93 , corresponding to sound rays that `` skim '' the horizon . on the other hand , when the trapping horizon consists of just one single sonic point located at @xmath94 , it is not obvious that an event horizon exists . loosely speaking ,
in this case it might happen that the trapping horizon form `` after '' every right - going ray from @xmath53 has managed to cross @xmath22 . since there is a competition between two infinite quantities the time at which the trapping horizon forms , and the time at which the `` last '' right - going signal that connects @xmath95 with @xmath96 crosses @xmath22 a careful case - by - case analysis is in order .
this is essentially all that can be said without relying on specific features of @xmath16 .
we now consider separately the various situations of interest , focussing first on the issue of the existence of the event horizon . in the case of a non - extremal black hole ,
the qualitative behaviour of the function @xmath16 is shown , graphically , in figure [ f : v - bh ] .
note that , for small values of @xmath97 , one can write@xmath98the function @xmath43 behaves as already shown in figure [ f : xi - bh ] .
a sketch of the worldlines of right - moving sound rays is presented in figure [ f : tx - bh ] .
note that in the portion of the diagram to the right of the curve @xmath67 ( , to the right of the moving kink in the velocity profile ) , spacetime is static . for @xmath52 ,
the geometry is minkowskian and the worldlines tend to approach straight lines with slope @xmath99 .
the sound ray that generates the event horizon corresponds to a finite ) , given the asymptotic behaviours of @xmath100 and @xmath30 . ]
value @xmath101 of the coordinate @xmath68 .
hence , in this situation an event horizon always exists .
this is also clear from the fact that the vertical half - line @xmath22 , @xmath102 in figure [ f : tx - bh ] is an apparent horizon .
the function @xmath16 behaves as shown in figure [ f : v - cbh ] .
regarding the right side of the profile , @xmath21 , it is indistinguishable from the profile of a non - extremal black hole ( figure [ f : v - bh ] ) . when the function @xmath43 is of the form ( [ appxi - bh ] ) , that is , when the apparent horizon is formed at a finite amount of laboratory time , the situation is exactly the same as for the non - extremal black hole discussed above .
consider now that the sonic point is approached in an infinite amount of time , so the function @xmath43 behaves as in figure [ f : xi - cbh ] .
the worldlines of right - moving sound rays are shown in figure [ f : tx - cbh ] .
as in the formation of the non - critical black hole , the portion of the diagram to the right of the curve @xmath67 ( , to the right of the moving kink in the velocity profile ) corresponds to a static spacetime , and for @xmath52 the geometry is minkowskian the worldlines tend to approach straight lines with slope @xmath99 .
however , now the apparent horizon is just the asymptotic point located at @xmath22 , @xmath4 , and in order to establish whether an event horizon does , or does not , exist one must perform an actual calculation of @xmath101 for the `` last '' ray that crosses the kink .
the expression for @xmath101 is again obtained from equation ( [ uu ] ) , noticing that now @xmath103 along the generator of the would - be horizon , so@xmath104the necessary and sufficient condition for the event horizon to exist is that the limit on right hand side of equation ( [ baru - cbh ] ) be finite .
the integrand on right hand side of ( [ baru - cbh ] ) can be approximated , for @xmath105 , as @xmath106 , while for @xmath52 it just approaches zero .
hence @xmath101 is , up to a finite constant , equal to @xmath107 times the integral of @xmath30 , evaluated at @xmath108 . here
we must distinguish between the exponential behaviour and the power law cases ( i ) and ( ii ) .
in the former @xmath101 is finite , trivially . for the power law
, it turns out that @xmath101 is finite iff @xmath109 .
the typical spatial profile function @xmath16 for an extremal black hole is plotted in figure [ f : v - ebh ] . for @xmath12
approaching zero from positive values we can write@xmath110where @xmath111 is a constant .
as far as dynamics is concerned , we must distinguish the cases in which the apparent horizon is formed at finite laboratory time @xmath34 , and in an infinite time ( , for @xmath4 ) .
the function @xmath43 is of the type shown in figure [ f : xi - bh ] , and the worldlines of right - going sound rays are sketched in figure [ f : rays - ex - fin ] .
the event horizon always exists .
the function @xmath43 is as shown in figure [ f : xi - cbh ] , and the worldlines of right - going signals are shown in figure [ f : tx - cbh ] .
as in the case of the formation of a critical black hole , the apparent horizon forms only asymptotically , for @xmath22 and @xmath4 , so the event horizon exists iff @xmath101 , given by equation ( [ baru - cbh ] ) , has a finite value . using the expansion ( [ appv - ebh ] ) in equation ( [ baru - cbh ] )
, one finds that @xmath112 is always finite when @xmath43 is asymptotically exponential . on the other hand , for a power law , the event horizon exists iff @xmath113 .
( note that the critical value of the exponent , @xmath114 , is now _ not _ the same as for the critical black hole , @xmath115 . )
the configurations we have analyzed until now are the simplest from a purely mathematical point of view . however , having in mind acoustic analogue geometries reproducible in a one - dimensional pipe in the laboratory , it is more sensible to consider double - sided configurations . by this
we mean that , after passing ( or approaching ) the sonic / supersonic regime at @xmath22 , and traversing an interval of width @xmath116 , the fluid again goes back to a subsonic regime as @xmath117 .
consider for example functions @xmath16 such that @xmath27 for @xmath118,@xmath119and which outside the interval @xmath118 tend monotonically to zero as @xmath97 increases ( see figures [ f : flat - spike ] and [ f : spike ] ) .
the corresponding fluid configuration represents what could be called a static `` double - sided critical black hole '' .
the formation of such a configuration can be modelled by the velocity function@xmath120with @xmath30 a monotonically decreasing function of @xmath13 , and @xmath100 as above .
accordingly , the differential equation for right - going sound rays also splits:@xmath121geometries associated with the formation of non - extremal and extremal black holes can be constructed in the same way ; see figures [ f : v - bh - wh ] and [ f : v - ex - bh ] for plots of the respective @xmath100 functions .
the function @xmath43 is of the type illustrated in figure [ f : xi - bh ] .
the behaviour of right - going sound rays is shown in figure [ f : rays - spike - fin ] .
the apparent horizon is the half - line @xmath22 , @xmath122 , and the event horizon always exists .
let us now consider a function @xmath43 of the type illustrated in figure [ f : xi - cbh ] .
the behaviour of right - going sound rays is shown in figure [ f : rays - spike - inf ] .
for these particular configurations , whether an event horizon does , or does not , actually exist is now a rather tricky issue .
the asymptotic behaviour of the function @xmath43 at @xmath123 ensures that all the right - going rays start to the left of the right - moving kink ( , the one in the region @xmath23 ) , then catch up with it , and begin to propagate through the intermediate region at a velocity @xmath124 that depends only on @xmath13 .
before they reach the point @xmath22 , such rays might be overtaken by the right - moving kink , but only to start the chase again .
after several mutual overtakings ( if the function @xmath43 is sufficiently complicated ) , the rays will always make an ultimate overtaking of the right - moving kink , embarking upon a final encounter with the left - moving kink on the right ( , in the region @xmath21 ) .
let us denote by @xmath125 the time of such a last crossing of the right - moving kink , so the corresponding event is @xmath126 .
also , let us denote by @xmath127 the time at which the same ray crosses the kink on the right , so that the corresponding event is @xmath128 .
from equation ( [ diffeq - dsc ] ) we directly obtain the relation@xmath129\;.\label{ziopollo}\]]between @xmath127 and @xmath125 .
( when the ray crosses the right - moving kink more than once , equation ( [ ziopollo ] ) will be satisfied by more than one value of @xmath125 for any given @xmath127 . in order to avoid cumbersome notation
, we shall simply denote by @xmath125 the largest of these roots , corresponding to the last crossing . )
then a necessary and sufficient condition for the existence of an event horizon is that , for @xmath130 , @xmath125 tends to a finite value , say @xmath131 .
this guarantees that any right - going ray that last crosses the left kink at a time greater than @xmath131 does not reach the region @xmath21 ( as ray - crossing can not occur under the working hypothesis of this paper ) . applying this condition straightforwardly in order to see whether the event horizon exists is not easy .
indeed , that would require us to evaluate the integral in equation ( [ ziopollo ] ) for a generic , finite value of @xmath127 , then solve for @xmath125 as a function of @xmath127 .
it is easier to use one of the following two alternative strategies : 1 . instead of asking whether the event horizon exists
, one can ask whether the event horizon _ does not _ exist .
a necessary and sufficient condition for this is that , for @xmath130 , also @xmath132 .
in such a case , we can insert the asymptotic expansions ( [ appv - bh ] ) or ( [ appv - ebh ] ) into equation ( [ ziopollo ] ) to get@xmath133for a non - extremal black hole , and@xmath134for an extremal one . plugging into these expressions the different asymptotic behaviours of the function @xmath43 , one can explicitly solve for @xmath125 as a function of @xmath127 for large values of the latter , and check whether @xmath125 does , or does not , tend to infinity when @xmath130 .
2 . setting @xmath135 into ( [ ziopollo ] ) , one obtains@xmath136\;.\label{xihinf}\]]it is possible to show that the event horizon exists if and only if equation ( [ xihinf ] ) possesses an odd number of finite solutions . is the solution of equation ( [ xihinf ] ) with the largest value . ] in order to establish whether this is the case , it is convenient to define the function of @xmath13@xmath137\;,\label{integral}\]]whose points of crossing with @xmath138 correspond to the solutions of equation ( [ xihinf ] ) . of course , for @xmath139 to be well defined ( and
therefore for solutions of ( [ xihinf ] ) to exist at all ) one needs the integral defining it to be convergent . for asymptotically ( @xmath140 ) exponential and power - law behaviours of @xmath43 this happens in the cases already described .
now , whenever @xmath139 is well defined , it is clearly a monotonically decreasing function , because the integrand in equation ( [ integral ] ) is always strictly positive . for @xmath123
, @xmath141 is just equal to the integral of the function @xmath124 evaluated at @xmath13 , up to a finite constant . in this limit , @xmath142 so we can write , for @xmath52:@xmath143given the condition @xmath144 , it is clear that the function @xmath141 is always greater than @xmath138 for @xmath52 . on the other hand , for @xmath145 ,
the asymptotic behaviour of @xmath141 is obtained by expanding @xmath100 in ( [ integral ] ) , which gives@xmath146for critical ( and non - extremal ) black holes and@xmath147for extremal ones .
if , for @xmath4 , @xmath141 is smaller ( greater ) than @xmath138 , then equation ( [ xihinf ] ) has an odd ( even ) number of finite solution , and the event horizon does ( does not ) exist .
note that , if @xmath148 , one must analyse subdominant terms in the asymptotic behaviour of @xmath43 in order to draw any conclusion . with either method , we find that for @xmath149 the existence of an event horizon in double - sided configurations follows the same rules as in the previously analysed one - sided configurations .
when @xmath150 however , it is more difficult to have an event horizon in a double sided configuration , and in general , one has to increase the _ rapidity _ with which one approaches the sonic regime .
more specifically , for a critical black hole and an asymptotically exponential @xmath43 , the event horizon exists if @xmath151 , but not when @xmath152 , while for a power law there is no horizon . for an extremal black hole and an exponential @xmath43 the horizon always exists , but in the case of a power law it does not exist if @xmath153 , and it exists for @xmath154 , with the additional condition @xmath155 for the particular value @xmath115 .
for a critical black hole with asymptotically exponential @xmath43 and @xmath156 , as well as for an extremal black hole with a power law and @xmath115 , @xmath157 , the asymptotic analysis is not sufficient and one must take into account also subdominant terms in the expansion of @xmath43 for @xmath145 .
for those situations in which an event horizon exists , we now find the asymptotic relation between @xmath69 and @xmath68 for rays close to the horizon generator .
we also briefly discuss the implications of such a relation for quasi - particle creation in the various cases of interest .
consider a sound ray corresponding to a value @xmath91 . for @xmath68 very close to @xmath101
, @xmath75 is very close to @xmath34 , and we can use the approximation ( [ appxi - bh ] ) for @xmath43 .
furthermore , we can approximate @xmath100 as in ( [ appv - bh ] ) , so equation ( [ u - fin ] ) gives@xmath158 ^ 2)\;.\label{ut0-bh}\]]this provides us with the link between @xmath68 and @xmath75 . in order to link @xmath75 with @xmath69 ,
consider the integral on the right hand side of equation ( [ uu ] ) .
for @xmath19 , the integrand function vanishes , while near @xmath159 it can be approximated by @xmath160
. then the integral is just given by the difference of the corresponding integrals evaluated at @xmath96 and @xmath161 , respectively , up to a possible finite positive constant .
this gives given in reference @xcite .
using equation ( 4.2 ) from that paper we have@xmath162expanding , we find again equation ( [ porcoqua ] ) , to the leading order in @xmath163.]@xmath164together with equation ( [ ut0-bh ] ) , this leads to@xmath165this relation between @xmath68 and @xmath69 is exactly the one found by hawking in his famous analysis of particle creation by a collapsing star @xcite .
it is by now a standard result that this relation implies the stationary creation of particles with a planckian spectrum at temperature @xmath166 @xcite . for a critical black hole ,
the results are very different according to whether the sonic regime is attained in a finite or an infinite laboratory time .
the calculation of the relation between @xmath68 and @xmath69 is exactly equal to the one presented for the non - extremal black hole case .
the two geometries coincide everywhere to the right of the apparent horizon and can not be distinguished by the quasi - particle production observed at @xmath167 .
let us suppose that we are in a situation in which the event horizon exists , so @xmath101 is finite . for another right - moving sound ray that corresponds to a value @xmath91
we find , combining equations ( [ uu ] ) and ( [ baru - cbh]),@xmath168\;.\label{deltau - cbh}\]]in the integration interval , @xmath43 is close to zero , so equation ( [ deltau - cbh ] ) can be approximated as@xmath169where the expansion ( [ appv - bh ] ) has been used .
equation ( [ deltau-cbh ] ) gives@xmath170for an asymptotically exponential @xmath30 , and@xmath171for a power law with @xmath109 . for the link between @xmath75 and @xmath69
we obtain@xmath172as one can easily check inserting the appropriate asymptotic expansions into equation ( [ uu ] ) . )
which , expanded , gives equation ( [ tu ] ) .
the result holds , however , independently of the details of @xmath16 .
] using equation ( [ tu ] ) into equations ( [ deltauexp ] ) and ( [ deltaunu ] ) we find@xmath173for the exponential case , and@xmath174for a power law with @xmath109 .
( remember that for @xmath175 the event horizon does not form . )
it is interesting to compare equations ( [ uuexp ] ) and ( [ uunu ] ) with the corresponding one for the non - critical black hole , equation ( [ uu-bh ] ) .
whereas the latter is basically independent of the details of the black hole formation ( which only appear in the multiplicative constant ) , the relation between @xmath68 and @xmath69 in the critical case is not universal , but depends on the dynamical evolution .
even for an asymptotically exponential @xmath43 , which leads to an exponential dependence on @xmath69 , the coefficient in the exponent is not universal as in equation ( [ uu-bh ] ) , but depends on dynamics through the parameter @xmath46 .
this is not difficult to understand looking back at the way in which equations ( [ uu-bh ] ) and ( [ uuexp ] ) have been derived . for equation ( [ uu-bh ] ) , the exponential dependence was introduced relating @xmath75 with @xmath69 , which only involves sound propagation in the final static region and can not , therefore , be affected by dynamics . on the other hand ,
when deriving equation ( [ uuexp ] ) it is sound propagation in the initial , dynamical , regime that introduces the exponential ( in the particular case of an asymptotically exponential @xmath30 ) ; hence , it is not surprising that the final result keeps track of the dynamical evolution .
however , it is interesting to note that in the limit @xmath176 equations ( [ uu-bh ] ) and ( [ uuexp ] ) coincide .
this limit corresponds to a very rapid approach towards the formation of an otherwise - never - formed ( in finite time ) apparent horizon .
regarding the creation of quasi - particles , this situation is operationally indistinguishable from the actual formation of the sonic point .
however , this `` degeneracy '' might be accidental , given that the origin of the exponential relation is very different in the two cases .
as for the case of a critical black hole , we must distinguish between a finite and an infinite time of formation of the event horizon .
for a sound ray close to the one that generates the horizon , equation ( [ u - fin ] ) still holds . however , now one must use the expansion ( [ appv - ebh ] ) when approximating the integrand thus obtaining@xmath177
^ 3)\;.\label{deltaut0-ebh1}\]]using again the approximation ( [ appv - ebh ] ) in the evaluation of the integral on the right hand side of equation ( [ uu ] ) one finds@xmath178finally,@xmath179interestingly , this is the same relation that one finds for the gravitational case @xcite . in particular
, this implies that finite time collapse to form an extremal black hole will _ not _ result in a planckian spectrum of quasi - particles @xcite .
this is completely compatible with the standard gr analysis , and is one of the reasons why extremal and non - extremal black holes are commonly interpreted as belonging to completely different thermodynamic sectors @xcite .
assuming that the event horizon exists , we can again apply equation ( [ deltau - cbh ] ) and use the approximation ( [ appv - ebh ] ) in order to find the relation between @xmath68 and @xmath75 .
the results are , for an asymptotically exponential @xmath43:@xmath180for a power law with @xmath181:@xmath182for a power law with @xmath115:@xmath183for a power law with @xmath109:@xmath184using the appropriate expansions in equation ( [ t - t ] ) , one obtains the relation between @xmath75 and @xmath69:@xmath185for an asymptotically exponential @xmath43 this becomes@xmath186for a power law , one must again distinguish between three cases ; for @xmath181:@xmath187for @xmath115:@xmath188for @xmath109:@xmath189putting together equations ( [ utexp - ebh ] ) and ( [ tuexp - ebh ] ) one finds the relationship between @xmath68 and @xmath69 for the exponential case:@xmath190for the power law one finds from equations ( [ utnu<1])([utnu>1 ] ) and ( [ utnu<1])([utnu>1 ] ) , for @xmath181:@xmath191for @xmath115:@xmath192and finally , for @xmath109:@xmath193 in all these cases , quasi - particle production is neither universal , nor planckian .
it is not difficult to prove that in the formation , in a finite amount of time , of double - sided non - extremal black holes , double - sided extremal black holes , and double - sided critical black holes , the asymptotic relation between @xmath68 and @xmath69 is identical to that calculated in the corresponding subsections above .
the amount and features of quasi - particle creation are then the same .
we will demonstrate this in detail for the case of a double - sided critical black hole , and then proceed to consider the situation in which the formation takes place in an infinite amount of time .
using the same notation as in section [ paragraph:2cbh2 ] , let us call @xmath125 the largest of the @xmath125 s that satisfy equation ( [ ziopollo ] ) , so @xmath125 is the time at which a right - going ray last crosses the kink on the left .
there will be some regular relationship between @xmath68 and @xmath125 , expressed by some differentiable function @xmath139 , so that we can write @xmath194 .
for the event horizon to exist , the corresponding @xmath68 must be finite ( equal to some value @xmath101 , say ) , so also @xmath195 must be finite ( as already done in section [ paragraph:2cbh2 ] , we denote by a suffix `` h '' the quantities that correspond to the horizon generator ) .
for a ray very close to the horizon generator we have@xmath196where a dot denotes the derivative with respect to @xmath13 . on the horizon , @xmath197 so equation ( [ ziopollo ] ) reduces to@xmath198\;.\label{ziogallo}\]]subtracting ( [ ziogallo ] ) from ( [ ziopollo ] ) we obtain@xmath199 -\int_{t_2}^{t_{\rm h}}\d t\left[c + \bar{v}\left(\xi(t)\right)\right]\;.\label{ziopollogallo}\]]for a ray close to the horizon generator , @xmath127 is close to @xmath200 , and @xmath125 close to @xmath131 , so equation ( [ ziopollogallo ] ) gives , keeping only terms to the leading order:@xmath201together , equations ( [ uf ] ) and ( [ eccola ] ) provide a linear link between @xmath68 and @xmath127 . since the relationship between @xmath127 and @xmath69 is exactly the same as the one between @xmath75 and @xmath69 in equation ( [ porcoqua ] ) ,
the final result is again the one expressed by ( [ uu-bh ] ) : @xmath202 assuming that we are in a situation for which the event horizon does indeed exist , we can subtract equation ( [ xihinf ] ) with @xmath203 from equation ( [ ziopollo ] ) , finding:@xmath204-\int_{t_2}^{+\infty}\d t\left[c + \bar{v}\left(\xi(t)\right)\right]\;.\label{xi - xi}\]]for a ray close to the horizon generator , @xmath125 is close to @xmath131 and @xmath127 is large , so@xmath205for an asymptotically exponential @xmath43 we find , performing the integral,@xmath206similarly , for a power law with @xmath109:@xmath207 in both cases , the same results as in section [ subsec : cbh ] , equations ( [ uuexp ] ) and ( [ uunu ] ) , follow . in short , the amount and characteristics of the quasi - particle production calculated with the double - sided configurations
are exactly the same as those calculated with the simpler profiles in the previous subsections except in two specific situations : the double - sided critical black hole with @xmath150 ( see figure [ f : spike ] ) and the double - sided extremal black hole . in the critical case
, only the asymptotically exponential behaviour with @xmath208 produces an event horizon and , therefore , only then we can talk about a stationary and planckian creation of quasi - particles . in the extremal case
the results described in section [ subsubsec : ebh2 ] only apply for @xmath209 ( with the further condition @xmath210 in the particular case @xmath115 ) , because otherwise the event horizon itself does not exist .
in the present paper we have analyzed different dynamical black hole - like analogue geometries with regard to their properties in terms of quantum quasi - particle production .
we have taken several @xmath1-dimensional spacetimes ( considered as externally fixed backgrounds ) , and for each of them we ( i ) have calculated whether it possesses an event horizon or not , and if the answer is `` yes '' , ( ii ) have calculated the asymptotic redshift function that characterizes the amount and properties of the late - time quasi - particle production . in table
[ table ] the reader can find a summary of all our results . .this is a summary of the results found for the different configurations analyzed in the paper . to the table we have to add the following comments : for double - sided extremal black hole with @xmath150 , with an infinite time of formation and an asymptotic power law for @xmath43 , the horizon forms in the case @xmath115 if the further condition @xmath155 holds . for the double - sided configurations with @xmath150 and an infinite time of formation , the asymptotic analysis is not sufficient for drawing conclusions when @xmath43 is asymptotically exponential and @xmath156 , and when @xmath43 is asymptotically a power law and @xmath115 , @xmath157 . for those cases ,
one need also consider subdominant terms in @xmath43 , so the results will depend on the details of formation . [ cols="^,^,^,^,^,^",options="header " , ]
the above results are pertinent to a purely mathematical model .
their physical relevance has to be assessed with respect to their application to both experimental reproduction of the analogue hawking radiation , and to the lessons they can provide concerning the possible behavior of black hole formation and evaporation in semiclassical gravity .
we now turn to separately consider these two issues .
the study carried on in this paper has identified several velocity profiles that are potentially interesting for experiments . in particular
the critical black hole models seem worth taking into consideration in connection with the realizability of a hawking - like flux in the laboratory . the creation of supersonic configurations in a laboratory is usually associated with the development of instabilities .
there are many examples of the latter in the literature ; in reference @xcite it was shown that in an analogue model based on ripplons on the interface between two different sliding superfluids ( for instance , @xmath211he - phase a and @xmath211he - phase b ) , the formation of an ergoregion would make the ripplons acquire an amplification factor that eventually would destroy the configuration .
therefore , this analogue system , although very interesting in its own right , will prove to be useless in terms of detecting a hawking - like flux .
however , by creating , instead of an ergoregion , a critical configuration one should be able , at least , to have a better control of the incipient instability , while at the same time producing a dynamically controllable hawking - like flux . nevertheless , the actual realization of a critical configuration could also appear as a problematic task for entirely different reasons .
the corresponding velocity profiles are characterized by discontinuities in the derivatives , so one might wonder whether they would be amenable to experimental construction , given that the continuum model is only an approximation .
let us therefore discuss in some detail the validity of the latter for realistic systems .
the main difference between an ideal perfect fluid model and a realistic condensed matter analogue is due to the microscopic structure of the system considered . in particular , it is generic to have a length scale @xmath212 which characterizes the breakdown of the continuum model ( @xmath212 is of the order of the intermolecular distance for an ordinary fluid ; of the coherence length for a superfluid ; and of the healing length for a bose einstein condensate ) . in general , the viability of the analogue model requires one to consider distances @xmath213 of order of at least a few @xmath212 , depending on the accuracy of the experiment performed . in particular ,
wave propagation is well defined only for wavelengths larger than @xmath212 ( generally with an intermediate regime , for wavelengths between @xmath212 and @xmath213 , where the phenomena exhibit deviations with respect to the predictions based on the continuum model ) . in general , a mathematical description based on the continuum model contains details involving scales smaller than @xmath213 ( for example , in the velocity profile ) .
these details should , however , be regarded as unphysical : they are present in the model , but do not correspond to properties of the real physical system . in particular
, they can not be detected experimentally , because this would require using wavelengths smaller than @xmath213 , which do not behave according to the predictions of the model ( and for wavelengths smaller than @xmath212 do not even make physical sense ) .
for the mathematical models considered in the present paper , all this implies that one will not be able to distinguish , on empirical grounds , between those cases for which the velocity profiles differ from each other only by small - scale details
. in particular , double - sided configurations with @xmath150 should be equivalent to configurations with a small , but non - zero , thickness @xmath214 .
also , one would not be able to distinguish between two velocity profiles that differ only in a neighborhood @xmath213 of @xmath22 , one of which corresponds to a critical black hole , while the other describes an extremal one . in particular hawking radiation
will not distinguish between the models within each of these pairs .
this fact would not be troublesome , had our analysis led to identical results for the acoustic black holes of each pair .
however , this is not the case ( see table [ table ] ) .
but then what shall we see if we realize these models in a laboratory ?
in realistic situations , what is relevant for hawking radiation is a coarse - grained profile obtained by averaging over a scale of order @xmath213 , thus neglecting the unphysical small scale details in @xmath16 .
this implies that as far as double - sided critical black holes are concerned , the reliable results are those pertinent to the non - zero thickness case ( @xmath215 ) .
similarly , since these extremal black holes are never exactly realizable in a laboratory ( as this would require tuning the velocity profile on arbitrary small scales ) , only the predictions based on the critical black hole mathematical model will survive in an experimental setting .
indeed , the relevant surface gravity will be defined by averaging the slope of the velocity profile over scales which are of order of @xmath213
. this averaged surface gravity will be non - zero for both the critical and the extremal black hole , but will be approximately equal to the surface gravity at the horizon of the critical black hole , while it will obviously not coincide with the one of the extremal ( which is zero ) . in the body of the paper
we have used a terminology particularly suitable to dealing with analogue models based on acoustics .
let us now discuss the most relevant features of our findings using a language more natural to gr .
when the geometry associated with the formation of a spherically symmetric black hole through classical gravitational collapse ( as , for example , in the oppenheimer - snyder model @xcite ) is described in terms of painlev - gullstrand @xcite coordinates ( whose counterpart , in the context of acoustic geometries , are the natural laboratory coordinates @xmath12 and @xmath13 ) , the apparent horizon forms in a finite amount of coordinate time . in this regard
, the painlev - gullstrand time behaves similarly to the proper time measured by a freely - falling observer attached to the surface of the collapsing star .
the non - extremal , non - critical @xmath1-dimensional model analysed in this paper , captures the main features of the formation of a ( non - extremal ) black hole .
the dynamical collapse is represented by the function @xmath43 in our calculations . in the language of gr
, we can think of @xmath43 as the radial distance between the surface of a collapsing star and its schwarzschild radius ; @xmath35 corresponds to the moment in which the surface of the star enters its schwarzschild radius , and this moment corresponds to a finite time ( which we took to be @xmath34 ) . for this model
we recovered hawking s result that the formation of ( non - extremal ) black holes causes the quantum emission towards infinity of a stationary stream of radiation with a planckian spectrum , at temperature @xmath166 .
the mechanism for particle creation is somewhat `` more than dynamical '' as the characteristics of the stationary stream of particles are `` universal '' and only depend on the properties of the geometry at the horizon , @xmath47 , and not on any detail of the dynamical collapse . indeed , for @xmath43 given by equation ( [ appxi - bh ] )
apparent horizon formation in a finite amount of time we have seen that asymptotic quasi - particle creation does not depend even on the coefficient @xmath38 .
that is , particle production does not depend on the velocity with which the surface of the collapsing star enters its schwarzschild radius .
this picture leans toward the ( quite standard ) view that hawking s process is not just dynamical , but relies on the actual existence of an apparent horizon and an `` ergoregion '' beyond it , able to absorb the negative energy pairs @xcite .
however , by analyzing alternative models , in this paper we have seen two unexpected things : i ) : : one can also produce a truly hawking flux with a temperature @xmath166 through the formation in a finite amount of time of either a single - sided critical black hole , or a double - sided critical black hole of finite thickness " , or even one of zero thickness " ( see figure [ f : spike ] ) .
this is an intriguing result , as in none of these cases there is an `` ergoregion '' beyond the apparent horizon , and in the last case there is just a single sonic point .
( in the language of gr , this last configuration corresponds to stopping the collapse of a star at the very moment in which its surface reaches the schwarzschild radius . )
ii ) : : moreover , one can also produce a stationary and planckian emission of quasi - particles by , instead of actually forming the apparent horizon , just approaching its formation asymptotically in time with sufficient rapidity ( @xmath216 ) . in this case
the temperature is not @xmath166 but @xmath217 , with@xmath218and ( at any finite time ) there is neither an apparent horizon nor an ergoregion within the configuration .
explanations of particle production based on tunneling then seem not viable , and the phenomenon is closer to being interpreted as dynamical in origin .
if fact , these configurations interpolate between situations in which the dynamics appears more prominently when @xmath219 we have that the temperature goes as @xmath220 and others in which the characteristics of the approached configuration are the more relevant and `` universality '' is recovered when @xmath221 we have that the temperature goes as @xmath166 , indistinguishable from hawking s result . by looking at our simple critical model
, we can say that , in geometrical ( kinematical ) terms , in order to obtain a steady and universal flux of particles from a collapsing ( spherically symmetric ) star there is no need for its surface to actually cross the schwarzschild radius ; it is sufficient that it tend towards it asymptotically ( in proper time ) , with sufficient rapidity .
our critical configurations could prove to be relevant also in the overall picture of semiclassical collapse and evaporation of black hole - like objects .
our results based on critical configurations suggest an alternative scenario to the standard paradigm . at this stage
we are only able to present it in qualitative and somewhat speculative terms .
being aware of the various assumptions that could ultimately prove to be untenable , we still think it is worth to present this possible alternative scenario .
imagine a dynamically collapsing star .
the collapse process starts to create particles dynamically before the surface of the star crosses its schwarzschild radius ( this particle creation is normally associated with a transient regime and has nothing to do with hawking s planckian radiation ) .
the energy extracted from the star in this way will make ( due to energy conservation ) its total mass decrease , and so also its schwarzschild radius . by this argument alone
, we can see that a process is established in which the surface of the star starts to closely chase its schwarzschild radius while both collapse towards zero ( this situation was already described by boulware in reference @xcite ) .
now , the question is : will the surface of the shrinking star capture its shrinking schwarzschild radius in a finite amount of proper time ?
let us rephrase this question in the language of this paper .
in an evaporating situation our function @xmath43 still represents the distance between the surface of the star and its schwarzschild radius .
the standard answer to the previous question is that @xmath43 becomes zero in a finite amount of proper time . to our knowledge , this view ( while certainly plausible ) is not guaranteed by explicit systematic and compelling calculations but still relies on somewhat qualitative arguments .
the standard reasoning can be presented as follows : for sufficiently massive collapsing objects , the classical behaviour of the geometry should dominate any quantum back - reaction at any and all stages of the collapse process , as hawking s temperature ( considered as an estimate of the strength of this back - reaction ) is very low ; quantum effects would be expected to become important only at the last stages of the evaporation process . however , in opposition to this standard view , stephens , t hooft , and whiting @xcite have argued for the mutual incompatibility of the existence of external observers measuring a hawking flux and , at the same time , the existence of infalling observers describing magnitudes beyond the apparent horizon .
the reason is that the operators describing any feature of the hawking flux do not commute ( and this non - commutation blows up at late times ) with the infalling components of the energy - momentum tensor operator at the horizon .
therefore , if we accept this argument , the presence of a hawking flux at infinity would be incompatible with the actual formation of the trapping horizon , which would be destroyed by the back - reaction associated to hawking particles .
this fact leads these authors ( seeking for self consistency ) , to look for the existence of a hawking flux ( or at least a flux looking very much like it ) in background geometries in which the collapse process of the star is halted , just before crossing the schwarzschild radius , producing a bounce .
( in our language this could be represented by a function @xmath43 monotonically decreasing from @xmath222 to some @xmath223 , at which it reaches a very small positive value , and then monotonically increasing from @xmath223 to @xmath224 . ) in their analysis they found exactly that : an approximately planckian spectrum of particles present at infinity during a sufficiently long time interval .
however , the modified behaviour that deviates the least from the classical collapse picture , and at the same time eliminates the trapping horizon , is that in which @xmath43 does not reach zero , but just `` asymptotically approaches zero '' at infinite proper time , and does that very quickly .
this is represented in our critical configurations by the exponential behaviour @xmath225 with a very large @xmath46 .
the interesting point is that the analysis in this paper suggests that with quasi - stationary configurations like this , one could expect quasi - stationary planckian radiation at a temperature very close to @xmath166 , just like in the hawking process .
standard gr suggests that the surface gravity @xmath47 ( inversely proportional to the total mass of the star ) would increase with time through the back - reaction caused by the quantum dissipation .
moreover , it is sensible to think that during the evaporation process @xmath46 would also depend on @xmath13 . as the evaporation temperature increases ( @xmath47 increases )
the back - reaction would become more efficient and therefore we might expect that @xmath46 decreases .
then , one could arrive at a situation as the one portrayed in figure [ f : critical - collapse ] .
the evolution of the evaporation temperature would interpolate between a starting temperature completely controlled by @xmath47 and a late time temperature completely controlled by @xmath46 , showing a possible semiclassical mechanism for regularizing the end point of the evaporation process .
in this scenario the complete semiclassical geometry will have neither an apparent horizon nor an event horizon .
in this circumstance there would be no trans - planckian problem , nor information loss associated with the collapse and evaporation of this black hole - like object .
whether this scenario is viable or not will be the subject of future work .
let us end by making a brief comment concerning modified dispersion relations .
everything we said in this paper assumes strict adherence to lorentz symmetry .
even if semiclassical gravity contained lorentz - violating traces in the form of modified dispersion relations at high energy , one would still expect that the resulting scenario for the collapse and evaporation of a black hole - like object would keep the quasi - stationary hawking - like flux of particles as a robust prediction @xcite .
however , the complete conceptual scenario could be very different . in the presence of dispersion at high energies ,
the notion of horizon itself shows up only as a low - energy concept .
for example , with superluminal modifications of the dispersion relations , high energy signals will be able to escape from the trapped region .
the non - analytic behaviour of some sets of modes at the horizon become regularized . therefore , the stephenst hooft whiting obstruction described above forbidding the formation of a ( now approximate ) trapping horizon need no longer apply . we expect that by analyzing different analogue models in which the lorentz violating terms appear at different energy scales one would be able to explore the transition between all these alternative paradigms .
the authors would like to thank vctor aldaya and ted jacobson for stimulating discussions .
. has been funded by the spanish mec under project fis2005 - 05736-c03 - 01 with a partial feder contribution .
c.b . and s.l
are also supported by a infn - mec collaboration .
the research of m.v . was funded in part by the marsden fund administered by the royal society of new zealand .
m.v . also wishes to thank both isas ( trieste ) and iaa ( granada ) for hospitality .
novello , m. visser and g. volovik ( eds . ) , _ artificial black holes _
( singapore , world scientific , 2002).c .
barcel , s. liberati and m. visser , `` analogue gravity , '' living rev .
relativity * 8 * , 12 ( 2005 ) [ arxiv : gr - qc/0505065 ] .
url ( cited on 22 march 2006 ) : http://www.livingreviews.org/lrr-2005-12c .
barcel , s. liberati and m. visser , `` analogue gravity from bose - einstein condensates , '' class .
quantum grav .
* 18 * , 11371156 ( 2001 ) [ arxiv : gr - qc/0011026].m .
visser , c. barcel and s. liberati , `` analogue models of and for gravity , '' gen .
relativ .
gravit .
* 34 * , 17191734 ( 2002 ) [ arxiv : gr - qc/0111111].c .
barcel , s. liberati , s. sonego and m. visser , `` causal structure of analogue spacetimes , '' new j. phys .
* 6 * , 186 ( 2004 ) [ arxiv : gr - qc/0408022].s .
w. hawking , `` black hole explosions , '' nature * 248 * , 3031 ( 1974 ) ; + `` particle creation by black holes , '' commun .
phys . * 43 * , 199220 ( 1975 ) ; erratum : _ ibid . _ * 46 * , 206 ( 1976 ) .
n. d. birrell and p. c. w. davies , _ quantum fields in curved space _
( cambridge , cambridge university press , 1982).m . visser , `` acoustic black holes : horizons , ergospheres , and hawking radiation , '' class .
quantum grav .
* 15 * , 17671791 ( 1998 ) [ arxiv : gr - qc/9712010].b .
l. hu , `` hawking - unruh thermal radiance as relativistic exponential scaling of quantum noise , '' in _ thermal field theory and applications _ , edited by y. x. gui , f. c. khanna and z. b. su ( singapore , world scientific , 1996 ) , pp .
249260 [ arxiv : gr - qc/9606073].m .
visser , `` essential and inessential features of hawking radiation , '' int .
j. mod .
d * 12 * , 649661 ( 2003 ) [ arxiv : hep - th/0106111].s .
w. hawking and g. f. r. ellis , _ the large scale structure of space - time _
( cambridge , cambridge university press , 1973).r .
m. wald , _ general relativity _
( chicago , university of chicago press , 1984).s . corley and t. jacobson , `` black hole lasers , '' phys . rev .
d * 59 * , 124011 ( 1999 ) [ arxiv : hep - th/9806203].r .
schtzhold and w. g. unruh , `` gravity wave analogues of black holes , '' phys .
d * 66 * , 044019 ( 2002).s .
liberati , t. rothman and s. sonego , `` nonthermal nature of incipient extremal black holes , '' phys .
d * 62 * , 024005 ( 2000 ) [ arxiv : gr - qc/0002019 ] .
s. liberati , t. rothman and s. sonego , `` extremal black holes and the limits of the third law , '' int .
j. mod .
d * 10 * , 3339 ( 2001 ) [ arxiv : gr - qc/0008018 ] .
j. r. oppenheimer and h. snyder , `` on continued gravitational contraction , '' phys .
rev . * 56 * , 455459 ( 1939).p .
painlev , `` la mcanique classique et la theorie de la relativit , '' c. r. acad .
( paris ) * 173 * , 677680 ( 1921 ) .
+ a. gullstrand , `` allgemeine lsung des statischen einkrperproblems in der einsteinschen gravitationstheorie , '' ark . mat .
* 16 * , 115 ( 1922).l .
parker , `` the production of elementary particles by strong gravitational fields , '' in _
asymptotic structure of space - time _ , edited by f. p. esposito and l. witten ( new york , plenum , 1977 ) , pp .
107226.g .
e. volovik , `` black - hole horizon and metric singularity at the brane separating two sliding superfluids , '' pisma zh .
fiz . * 76 * , 296300 ( 2002 ) [ jetp lett . *
76 * , 240244 ( 2002 ) ] [ arxiv : gr - qc/0208020].d .
g. boulware , `` hawking radiation and thin shells , '' phys .
d * 13 * , 21692187 ( 1976).a .
ashtekar and m. bojowald , `` black hole evaporation : a paradigm , '' class .
quantum grav .
* 22 * , 33493362 ( 2005 ) [ arxiv : gr - qc/0504029].c .
r. stephens , g. t hooft and b. f. whiting , `` black hole evaporation without information loss , '' class .
quantum grav .
* 11 * , 621647 ( 1994 ) [ arxiv : gr - qc/9310006].w .
g. unruh and r. schtzhold , `` on the universality of the hawking effect , '' arxiv : gr - qc/0408009 . | we discuss the issue of quasi - particle production by `` analogue black holes '' with particular attention to the possibility of reproducing hawking radiation in a laboratory . by constructing simple geometric acoustic models
, we obtain a somewhat unexpected result : we show that in order to obtain a stationary and planckian emission of quasi - particles , it is _ not _ necessary to create an ergoregion in the acoustic spacetime ( corresponding to a supersonic regime in the flow ) .
it is sufficient to set up a dynamically changing flow _ either _ eventually generating an arbitrarily small sonic region @xmath0 , but without any ergoregion , _ or _ even just asymptotically , in laboratory time , approaching a sonic regime with sufficient rapidity .
pacs : 04.20.gz , 04.62.+v , 04.70.-s , 04.70.dy , 04.80.cc + keywords : analogue models , acoustic spacetime , hawking radiation |
in spite of the considerable efforts to explain the experimental raman spectra of cuprate superconductors , the @xmath0 superconducting response is not yet completely understood . it has been shown that the theoretical description of the @xmath0 raman response was very sensitive to small changes in the raman vertex harmonic representations , yielding peak positions varying between @xmath1 and 2@xmath1 @xcite . however , the data show peaks consistently slightly above @xmath1 for both ybco and bscco . in this paper we present calculations suggesting that the @xmath0 peak position is largely controlled by a collective spin fluctuation ( sf ) mode near 41 mev , consistent with inelastic neutron scattering ( ins ) observations @xcite .
we show that the @xmath0 response is strongly modified by the sf term and is not sensitive to small changes in the raman vertex .
the experimental peak position is well reproduced by our model whereas the @xmath2 and @xmath3 response remain essentially unaffected by the sf mode .
the cuo@xmath4 bilayer is modeled by a tight binding band structure with a nearest ( @xmath5 ) and a next nearest neighbor hopping ( @xmath6 ) parameter and an inter - plane hopping given by @xcite @xmath7 ^ 2 .\ ] ] @xmath8 can be 0 or @xmath9 , for bonding or anti - bonding bands of the bilayer , respectively .
the spin susceptibility ( @xmath10 ) is modeled by extending the weak coupling form of a @xmath11 superconductor to include antiferromagnetic spin fluctuations by an rpa form with an effective interaction @xmath12 ; i.e. @xmath13 where @xmath14 is the simple bubble in the d - wave state .
this form of the spin susceptibility is motivated by the fact that it contains a strong magnetic resonance peak at @xmath15 which was proposed @xcite to explain the ins resonance at energies near 41 mev in ybco @xcite and bscco @xcite .
the raman response function in the superconducting state is evaluated using nambu green s functions .
the spin fluctuations contribute to the raman response via a 2-magnon process as shown in fig .
[ total ] @xcite where a schematic representation of the feynman diagrams of the sf and the bubble contribution is plotted . for the electronic propagators
we have used the bare bcs green s functions and a d - wave superconducting gap @xmath16/2 $ ] .
the total raman response is calculated in the gauge invariant form which results from taking into account the long wavelength fluctuations of the order parameter @xcite .
the total raman susceptibility is thus given by @xmath17 where @xmath18 is determined according to fig .
[ total ] . the analytical continuation to the real axis
is performed using pad approximants .
we have used several different forms for the raman vertex @xmath19 which possess the correct transformation properties required by symmetry .
our calculations show that the sf term yields vanishingly small corrections to the response in the @xmath2 and @xmath3 channels , but contributes substantially to the @xmath0 channel .
the shape of the total response in the @xmath0 geometry is mainly dependent on the value of the effective interaction @xmath12 .
variations of @xmath12 change the relative magnitude of the two diagrams summed in fig .
[ total ] , changing the position of the peak in @xmath0 geometry .
importantly , we find that the @xmath0 response shows little dependence on the form used for the vertex : @xmath20 , or the vertex calculated in an effective mass approximation .
these results can be explained by symmetry reasons given that the sf propagator is strongly peaked for @xmath21 momentum transfers .
we compare the calculated raman response with the experimental spectra of an optimally doped bi-2212 sample @xcite in fig .
[ fig22 ] . adding the sf contribution leads to a shift of the peak position from near @xmath22 for @xmath23 to higher frequencies , allowing a better agreement with the experimental relative positions of the peaks in @xmath0 and @xmath2 geometries . for the fit we have adjusted @xmath5 to achieve a good agreement with the @xmath2 channel , obtaining @xmath24 mev , and then adjusted @xmath12 to match both the @xmath0 peak position as well as the peak in the sf propagator to be consistent with the ins peak at 41 mev . from this work
we conclude that including the sf contribution in the raman response solves the previously unexplained sensitivity of the @xmath0 response to small changes in the raman vertex .
whereas the sf ( two - magnon ) contribution controls the @xmath0 peak , the @xmath2 and @xmath3 scattering geometries are essentially unaffected and determined by the bare bubble alone . 9 t.p .
et al . _ ,
phys . rev .
b * 51 * , 16336 ( 1995 ) ; _ ibid . _ * 54 * , 12523 ( 1996 ) .
et al . _ ,
lett . * 75 * , 316 ( 1995 ) ; nature * 398 * , 588 ( 1999 ) .
et al . _ , phys .
rev . lett . * 70 * , 3490 ( 1993 ) .
et al . _ ,
b * 53 * , 5149 ( 1996 ) .
a.p . kampf and w. brenig , z. phys .
b- condensed matter * 89 * , 313 ( 1992 ) .
rev . lett . * 72 * , 3291 ( 1994 ) . | while the low frequency electronic raman response in the superconducting state of the cuprates can be largely understood in terms of a d - wave energy gap , a long standing problem has been an explanation for the spectra observed in @xmath0 polarization orientations .
we present calculations which suggest that the peak position of the observed @xmath0 spectra is due to a collective spin fluctuation mode . |
graph theory has recently reveived increasing attraction for applications to complex systems in various disciplines ( gernert 1997 , paton 2002a , b , bornholdt and schuster 2003 ) .
the characterization of systems ( with interrelated constituents ) by graphs ( with linked vertices ) is comparably general as their characterization in terms of categories ( with elements related by morphisms ) . despite its generality ,
graph theory has turned out to be a powerful tool for gaining very specific insight into structural and dynamical properties of complex systems ( see jost and joy 2002 , atmanspacher et al .
2005 for examples ) .
an area of particularly intense interest , in which complex systems abound , is biological information processing .
this ranges from evolutionary biology over genetics to the study of neural systems .
theoretical and computational neuroscience have become rapidly growing fields ( hertz et al .
1991 , haykin 1999 , dayan and abbott 2001 ) in which graph theoretical methods have gained considerable significance ( cf .
sejnowski 2001 ) .
two basic classes of biological networks are feedforward and recurrent networks . in networks with purely feedforward ( directed ) connectivities , neuronal input is mapped onto neuronal output through a feedforward synaptic weight matrix . in recurrent networks , there are additional ( directed or bi - directed ) connectivities between outputs and other network elements , giving rise to a recurrent synaptic weight matrix . much recurrent modeling incorporates the theory of nonlinear and complex dynamical systems ( cf .
smolensky 1988 , see also beim graben 2004 for discussion ) .
hopfield networks are an example of a fully recurrent network in which all connectivities are bidirectional and the output is a deterministic function of the input .
their stochastic generalizations are known as boltzmann machines .
another important distinction with respect to the implementation of neural networks refers to the way in which the neuronal states are characterized : the two main options are firing rates and action potentials ( for more details see haykin 1999 ) . a key topic of information processing in complex biological networks is learning , for which three basically different scenarios are distinguished in the literature ( see dayan and abbott 2001 , chap .
iii ) : unsupervised , supervised and reinforcement learning . in unsupervised (
also self - supervised ) learning a network responds to inputs solely on the basis of its intrinsic structure and dynamics .
a network learns by evolving into a state that is constrained by its own properties and the given inputs , an important modelling strategy for implicit learning processes .
in contrast , supervised learning presupposes the definiton of desired input - output relations , so the learned state of the network is additionally constrained by its outputs .
usually , the learning process in this case develops by minimizing the difference between the actual output and the desired output .
the corresponding optimization procedure is not intrinsic to the evolution of the system itself , but has to be externally arranged , hence the learning is called supervised .
if the supervision is in some sense `` naturalized '' by coupling a network to an environment , which provides evaluative feedback , one speaks of reinforcement learning . in this contribution
we are interested in supervised learning ( see duda et al .
2000 for a review ) on small , fully recurrent networks implemented on graphs ( cf .
jordan 1998 ) .
we start with a general formal characterization in terms of dynamical systems ( sec .
2.1 ) , describe how they are implemented on graphs ( sec .
2.2 ) , and show how it reaches asymptotically stable states ( attractors ) when the learning process is terminated , i.e. is optimized for given inputs and ( random ) initial conditions with respect to predetermined outputs ( sec .
. we shall characterize the learning operations by a multiplicative structure characterizing successively presented inputs in sec . 3.1 . in this context
we confirm and specify earlier conjectures ( e.g. , gernert 1997 ) about the non - commutativity of learning operations for a concrete model . in sec .
3.2 , we study how the size of the set of attractors representing the derived structure changes during the process for perfectly and imperfectly optimized networks .
the number of attractors is proposed to indicate the complexity of learning , and in sec . 4 this
is tentatively related to pragmatic information as a particular measure of meaning .
let @xmath0 be a set , and let @xmath1 , with @xmath2 , be a partition of @xmath0 into two disjoint subsets .
if @xmath0 is some closed subset of @xmath3 , @xmath4 may be the boundary of @xmath0 .
( later we will specify @xmath0 as the vertices of a graph , @xmath4 as a set of `` external '' or `` boundary '' vertices , and @xmath5 as a set of `` internal '' vertices . )
we consider the dynamics of fields @xmath6 , where @xmath7 , @xmath8 , @xmath9 represents time as parametrized discretely or continuously , and @xmath10 is the space of admissible state values for the fields .
the dynamics of @xmath11 can be described by an equation @xmath12=0 \ , .\ ] ] for a continuous time variable and @xmath13 , a typical example is the diffusion equation @xmath14= \frac{\partial u(x , t)}{\partial t } - \lambda \delta u(x , t)\ ] ] where @xmath15 is the laplace operator and @xmath16 the diffusion constant . the only constraint on eq .
[ eq1 ] is that a state @xmath17 at time @xmath18 determines uniquely the solution for any time @xmath19 .
we now define a set of external conditions @xmath20 specifying field values @xmath21 on @xmath4 which will be kept fixed during the time evolution of the fields on @xmath0 . this is to say that the dynamics of fields is effectively restricted to @xmath5 : @xmath22=0\ , .\ ] ] since the state of the system at time @xmath18 uniquely determines the states for all @xmath19 , we can define a mapping @xmath23 , the so - called time evolution operator , acting on the set of field states . for an initial state @xmath11 at @xmath18 , @xmath24 $ ] yields the state of the system at @xmath19 . taking into account
that different external conditions initiate different evolutions , we have to specify the time evolution operator as a mapping @xmath25 , where @xmath26 is the set of states @xmath27 , by the following construction : let @xmath28 be the initial condition for eq .
( [ eq2 ] ) , then @xmath29(x ) = u(x , b_i , t)\ ] ] is the state of the corresponding solution at time @xmath9 under the external condition @xmath21 . in principle
, the state space of @xmath30 can be the entire set of states @xmath26 .
however , for reasons which will become clear below , we are interested in dissipative systems evolving into attractors @xmath31 in the limit of large @xmath9 .
if one of the states belonging to an attractor is chosen as an initial condition , the image of @xmath30 will again be one of the attractor states .
this allows us to reduce the number of possible states on which the mappings @xmath30 close . denoting the flow operator @xmath32 as the input under the external condition @xmath21
, we now consider the set of states @xmath33 belonging to attractors after time @xmath9 .
then all mappings @xmath34 , applied to an attractor @xmath35 , lead to images in @xmath36 : @xmath37 \in a \hspace{1 cm } \mbox{for } a\in a \ , .\ ] ] in general , the set of all attractor states @xmath36 does not contain a proper subset which is mapped onto itself by the set of mappings @xmath38 ; otherwise @xmath36 can be reduced to such a subset .
each single mapping @xmath34 may not be surjective , but the union of the images of all @xmath38 equals @xmath36 . due to condition ( [ eq3 ] ) , we can define a composition of mappings @xmath34 . in this way ,
the external conditions @xmath39 give rise to an associative multiplicative structure @xmath38 .
this structure is represented on the set of attractors @xmath40 .
we now implement the general notions developed so far on graphs ( see wilson 1985 for an introduction to graph theory ) and specify the set @xmath0 as the set of vertices @xmath41 of a graph . for simplicity
we consider directed graphs with single connections for each direction between any two vertices and without self - loops .
such a graph gives rise to non - reflexive relations on @xmath41 and can be represented by an adjacency matrix @xmath42 . for two vertices @xmath43 and @xmath44 we have : @xmath45 if @xmath42 is symmetric , the graph is undirected .
the set of vertices @xmath41 is decomposed into a set of external vertices @xmath46 and a set of internal vertices @xmath47 .
if @xmath48 is the total number of vertices , @xmath49 the number of external vertices and @xmath50 the number of internal vertices , we have @xmath51 .
next we consider fields @xmath52 on a graph with vertices @xmath53 evolving in discrete time steps @xmath54 according to : @xmath55 the value of the field @xmath11 at vertex @xmath56 and time @xmath57 depends only on the sum of the field values at neighboring vertices @xmath58 at time @xmath9 .
the fields @xmath52 assume integer values @xmath59 , and the function @xmath60 is defined as : @xmath61 where @xmath62 denotes the nearest integer - rounded @xmath63 .
the function @xmath64 is shown in fig .
[ fx ] .
( 240,100)(0,0 ) ( 10,10)(1,0)200 ( 10,10)(0,1)80 ( 60,8)(0,1)4 ( 160,8)(0,1)4 ( 8,60)(1,0)4 ( 15,15)(5,5)10 ( 70,55)(10,-5)10 ( 65,55)(10,-5)10 ( 165,10)(5,0)5 ( 215,5)(0,0)@xmath63 ( 60,3)(0,0)@xmath65 ( 160,3)(0,0)@xmath66 ( 0,60)(0,0)@xmath67 ( 5,100)(0,0)@xmath64 the restriction of @xmath52 to integer values implies that there is only a finite number of states . starting from an arbitrary initial state in @xmath26 , the system runs into an attractor after a few time steps .
in many cases , this attractor is a fixed point , i.e. one single state that is asymptotically stable .
sometimes the attractor is a limit cycle , i.e. a periodic succession of several ( usually few ) states .
strange attractors do not occur since the number of states is finite .
the external conditions @xmath39 are defined as fixed states on the external vertices , i.e. , the state values on the external vertices remain unaffected by the dynamics . of course , the external conditions are supposed to influence the dynamics of the internal vertices .
the graphs used in our investigations consist of a total of @xmath68 vertices with @xmath69 external vertices and @xmath70 internal vertices .
the maximal value of @xmath52 is defined to be @xmath71 .
we consider 11 different input patterns @xmath21 which are shown in fig .
[ fig1 ] .
( 390,120)(-20,0 ) ( 0,80 ) ( 10,80 ) ( 20,80 ) ( 30,80 ) ( 0,90 ) ( 10,90 ) ( 20,90 ) ( 30,90 ) ( 0,100 ) ( 10,100 ) ( 20,100 ) ( 30,100 ) ( 0,110 ) ( 10,110 ) ( 20,110 ) ( 30,110 ) ( 140,80 ) ( 150,80 ) ( 160,80 ) ( 170,80 ) ( 140,90 ) ( 150,90 ) ( 160,90 ) ( 170,90 ) ( 140,100 ) ( 150,100 ) ( 160,100 ) ( 170,100 ) ( 140,110 ) ( 150,110 ) ( 160,110 ) ( 170,110 ) ( 210,80 ) ( 220,80 ) ( 230,80 ) ( 240,80 ) ( 210,90 ) ( 220,90 ) ( 230,90 ) ( 240,90 ) ( 210,100 ) ( 220,100 ) ( 230,100 ) ( 240,100 ) ( 210,110 ) ( 220,110 ) ( 230,110 ) ( 240,110 ) ( 280,80 ) ( 290,80 ) ( 300,80 ) ( 310,80 ) ( 280,90 ) ( 290,90 ) ( 300,90 ) ( 310,90 ) ( 280,100 ) ( 290,100 ) ( 300,100 ) ( 310,100 ) ( 280,110 ) ( 290,110 ) ( 300,110 ) ( 310,110 ) ( 350,80 ) ( 360,80 ) ( 370,80 ) ( 380,80 ) ( 350,90 ) ( 360,90 ) ( 370,90 ) ( 380,90 ) ( 350,100 ) ( 360,100 ) ( 370,100 ) ( 380,100 ) ( 350,110 ) ( 360,110 ) ( 370,110 ) ( 380,110 ) ( 0,10 ) ( 10,10 ) ( 20,10 ) ( 30,10 ) ( 0,20 ) ( 10,20 ) ( 20,20 ) ( 30,20 ) ( 0,30 ) ( 10,30 ) ( 20,30 ) ( 30,30 ) ( 0,40 ) ( 10,40 ) ( 20,40 ) ( 30,40 ) ( 70,10 ) ( 80,10 ) ( 90,10 ) ( 100,10 ) ( 70,20 ) ( 80,20 ) ( 90,20 ) ( 100,20 ) ( 70,30 ) ( 80,30 ) ( 90,30 ) ( 100,30 ) ( 70,40 ) ( 80,40 ) ( 90,40 ) ( 100,40 ) ( 140,10 ) ( 150,10 ) ( 160,10 ) ( 170,10 ) ( 140,20 ) ( 150,20 ) ( 160,20 ) ( 170,20 ) ( 140,30 ) ( 150,30 ) ( 160,30 ) ( 170,30 ) ( 140,40 ) ( 150,40 ) ( 160,40 ) ( 170,40 ) ( 210,10 ) ( 220,10 ) ( 230,10 ) ( 240,10 ) ( 210,20 ) ( 220,20 ) ( 230,20 ) ( 240,20 ) ( 210,30 ) ( 220,30 ) ( 230,30 ) ( 240,30 ) ( 210,40 ) ( 220,40 ) ( 230,40 ) ( 240,40 ) ( 280,10 ) ( 290,10 ) ( 300,10 ) ( 310,10 ) ( 280,20 ) ( 290,20 ) ( 300,20 ) ( 310,20 ) ( 280,30 ) ( 290,30 ) ( 300,30 ) ( 310,30 ) ( 280,40 ) ( 290,40 ) ( 300,40 ) ( 310,40 ) ( 350,10 ) ( 360,10 ) ( 370,10 ) ( 380,10 ) ( 350,20 ) ( 360,20 ) ( 370,20 ) ( 380,20 ) ( 350,30 ) ( 360,30 ) ( 370,30 ) ( 380,30 ) ( 350,40 ) ( 360,40 ) ( 370,40 ) ( 380,40 ) ( 15,70)(0,0)1 ( 155,70)(0,0)2 ( 225,70)(0,0)3 ( 295,70)(0,0)4 ( 365,70)(0,0)5 ( 15,0)(0,0)6 ( 85,0)(0,0)7 ( 155,0)(0,0)8 ( 225,0)(0,0)9 ( 295,0)(0,0)10 ( 365,0)(0,0)11 .example of a mapping diagram for 11 inputs @xmath34 and a system with 10 different attractors @xmath31 .
the entries show the number @xmath72 of the attractor state which is obtained by applying @xmath34 ( plotted vertically ) to @xmath31 ( plotted horizontally ) . [ cols= " > ,
> , > , > , > , > , > , > , > , > , > " , ] the multiplicative structure associated with tab .
[ tab2 ] consists of the 11 elements @xmath34 which are idempotent , @xmath73 and satisfy the relation @xmath74 hence they are non - commutative , though associative : @xmath75 since the optimal reaction of a graph to an input is not uniquely related to that input , the attractor providing an optimal output can be identical for different inputs .
therefore , the multiplicative structure of input operations can be even simpler in the sense that some of the attractors are identical .
table [ tab0 ] shows a corresponding example with less than 11 attractors .
deviations from eq .
( [ eqproj ] ) indicate a more complicated structure of learning operations .
if the elements in the same row ( i.e. for the same input ) of the mapping diagram differ from each other , the reaction of the graph with respect to an input depends on the previous input .
this means that the result of a learning process depends on the sequence in which successive learning steps are carried out .
this implies that the multiplicative structure of input operations deviates from eq .
( [ eqproj ] ) .
since the @xmath34 are mappings , associativity is valid trivially .
however , the structure will generally be non - commutative , @xmath76 although it may happen that particular inputs commute , for instance when they project onto the same attractor , such as @xmath77 and @xmath78 , or @xmath79 and @xmath80 , or @xmath81 and @xmath82 in tab .
[ tab0 ] . we can now understand how an optimal learner differs from a perfect learner , which recognizes inputs independently of the sequence of their presentation . comparing tabs .
[ tab0a ] and [ tab1 ] shows that attractor @xmath83 leads to the optimal output ( field values on the first two vertices ) for input @xmath84 , attractors @xmath85 and @xmath86 yield the optimal output for inputs @xmath87 , and attractors @xmath88 and @xmath89 yield the optimal output for inputs @xmath90 . in these cases ,
optimal learning coincides with perfect learning . from tab .
[ tab0 ] we see that inputs @xmath91 are recognized independently of previous inputs .
by contrast , inputs @xmath92 , @xmath93 and @xmath94 are recognized correctly only if the previous input is @xmath95 , @xmath92 and @xmath93 , respectively .
table [ tab0a ] shows that attractors @xmath96 lead to an `` almost '' correct output for inputs @xmath97 , and the output of @xmath98 differs considerably from any optimal output .
although these situations represent optimal learning , they are different or even far from perfect learning . if the attractor for a particular input does not consist of one single state ( fixed point ) , but of a perodic sequence of states ( limit cycle ) , idempotency [ eqidemp ] does no longer hold .
( strictly speaking , this is only correct if the number of time steps @xmath9 in the mapping @xmath99 and the length of the cycle have no common denominator .
otherwise , the attractor may consist of more states than can be detected by the mapping diagram or the set of inputs @xmath34 . )
note that the structure of learning operations derived here is more general than an algebra ( as conjectured by gernert 2000 ) .
there is no identity element , there is no neutral element , and no addition of the elements @xmath34 is defined . in order to investigate the evolution of the set of attractors during the learning process
, we focus on the number @xmath48 of attractor states as a function of learning steps for the entire sequence of graphs starting from a random graph until a graph with optimal learning is reached .
since a large number of attractors intuitively relates to quite complex structures of the graph during the learning process , we propose to refer to the size of the set of attractors as a possible measure for the _ complexity of learning_. however , it should be emphasized that a rigorous definition of complexity ( cf .
wackerbauer et al .
1994 ) is not yet associated with this notion .
initially , the graphs are ( almost ) random and exhibit large variances of the order of @xmath100 . for these graphs
the number of attractor states with respect to the inputs varies over a large range ; typical are numbers between 30 and 50 .
as learning begins , the variance decreases , but the number of attractor states increases , sometimes up to a few hundred .
a further decrease in variance , below a value of 6000 , causes the number of attractor states to decrease again . for optimal learners ( graphs with vanishing variance ) the number of attractor states terminates at around @xmath101 .
a typical example is shown in fig .
[ fig3 ] .
we now select a sample of 116 learning sequences starting from random graphs and terminating as ( almost ) optimal learners . for this sample
we count the number of attractor states , i.e. the complexity of learning , for those graphs which were accepted during the process , i.e. , for which the variance was always smaller than for any previous graph in the sequence .
their behavior can be seen in fig .
[ fig4 ] , where @xmath48 is plotted as a function of @xmath102 .
it confirms the impression from fig .
[ fig3 ] that , as learning proceeds , its complexity evolves non - monotonically .
in about 50% of the cases the sequence started with less than @xmath103 attractor states .
the final @xmath48 was much smaller , and for intermediate stages of learning @xmath48 reached a maximum during the learning process . in about 85% of all cases
the final number of attractor configurations was smaller than 20 .
the largest final number of attractor states for an optimal learner was 56 .
exceptions from this behavior occur if the initial ( random ) graph has a number of attractor states that is extremely large , exceeding any other number of attractor states in the sequence .
for this case we find a total number of 15 sequences . in 12 of these sequences
the initial number of attractors is larger than 100 ( with a maximum of 747 ) .
figure [ fig5 ] shows a plot of number of attractors as a function of variance for 98 non - optimal learners whose final variance is @xmath104 . keeping in mind that decreasing variance corresponds to progressive learning , the general trend of figs .
[ fig3 ] and [ fig4 ] reappears : the size of the set of attractors , i.e. the complexity of learning , evolves non - monotonically as learning proceeds . as the main observation of the present subsection
, we can state that the number @xmath48 of attractors required to optimally map a given input onto a predetermined output evolves non - monotonically during the process of learning . while @xmath48 increases during the initial phase of learning , it decreases again until the learning process is terminated .
we interpret this behavior as a non - monotonic complexity of the learning process .
non - monotonic as opposed to monotonic measures of complexity have been developed and investigated for about two decades ; for a comparative overview see wackerbauer et al .
the property of monotonicity is usually understood as a function of ( some measure of ) randomness of the pattern or process considered .
monotonic complexity essentially increases as randomness increases : most random features are also most complex .
non - monotonic complexity shows convex behavior as a function of increasing randomness : highest complexity is assigned to features with a mixture of random and non - random elements , while both very low and very high randomness yield minimal complexity .
there is an interesting relationship between the two classes of complexity measures and measures of information ; for more details see atmanspacher ( 1994 ) or atmanspacher ( 2005 ) .
it turns out that monotonic complexity usually corresponds to syntactic information , whereas non - monotonic ( convex ) complexity corresponds to semantic information or other measures of meaning ( see fig .
[ fig6 ] ) . as a particularly interesting approach ,
pragmatic information has been proposed ( weizscker 1972 ) as an operationalized measure of meaning .
its essence is that purely random messages keep providing complete novelty ( or primordiality ) as they are delivered , while purely non - random messages keep providing complete confirmation ( after initial transients ) .
pragmatic information refers to meaning in terms of a mixture of confirmation and novelty . extracting meaning from a message depends on the capability to transform novel elements into knowledge using confirming elements .
it has been speculated ( atmanspacher 1994 ) that systems having this capacity are able to reorganize themselves in order to flexibly modify their complexity relative to the task that they are supposed to solve .
a learning process , in which insight is gained and meaning is understood , may start at low complexity ( high randomness , much novelty ) and terminate at low complexity ( high regularity , much confirmation ) , but it passes through an intermediate stage of maximal complexity .
the notion of pragmatic information was earlier utilized in this sense for non - equilibrium phase transitions in multimode lasers ( atmanspacher and scheingraber 1990 ) .
it could be shown that a particular well - defined type of pragmatic information , adapted to that case , behaves precisely as indicated above .
pragmatic information is maximal at the unstable stage of the phase transition , and it is low in the preceding and successive stages .
however , lasers are physical systems , and it is problematic to ascribe something like an `` understanding of meaning '' to their behavior .
biological networks such as studied in this paper are more realistic systems for a concrete demonstration of the basic idea .
the non - monotonic complexity of learning processes as indicated in sec .
3.2 starts with random graphs and ends with graphs of minimized variance ( maximized fitness ) , which are as non - random as possible under the given conditions . in this sense
, a scenario has been established in which the complexity of learning on graphs qualitatively satisfies the conditions required for relating it to a measure of pragmatic information . within this scenario ,
our approach suggests that the actual `` release of meaning '' during learning does not occur when the output is optimized but rather when the complexity is maximized .
it is a long - standing desideratum to identify meaning - related physiological features in the brain ( freeman 2003 ) .
since learning is a key paradigm in which the emergence of meaning can be studied , we hope that our approach may offer a useful perspective for progress concerning this problem .
in this contribution an example of supervised learning in recurrent networks of small size implemented on graphs is studied numerically .
the elements of the network are treated as vertices of graphs and the connections among the elements are treated as links of graphs .
eleven inputs and two outputs are predefined , and the learning process within the remaining six internal vertices is carried out such as to minimize the difference between the actual output and the predetermined output .
optimization of outputs is achieved by stable configurations at the internal vertices that can be characterized as attractors .
two particular features of the learning behavior of the network are investigated in detail .
first , it is shown that , in general , the mapping from inputs to outputs depends on the sequence of inputs .
thus , the associative multiplicative structure of input operations represented by sets of attractors is , in general , non - commutative .
second , the size of the set of attractors changes as the learning process evolves .
with increasing optimization ( fitness ) , the number of attractors increases up to a maximum and then decreases down to a usually small final set for optimal network performance . assuming that the size of the set of attractors indicates the complexity of learning , its non - monotonic behavior is of special interest .
since non - monotonic measures of complexity can be related to pragmatic information as a measure of meaning , it is tempting to consider the maximum of complexity as reflecting the release of meaning in learning processes .
further work will be necessary to substantiate this speculation .
atmanspacher , h. , 1994 , complexity and meaning as a bridge across the cartesian cut . journal of consciousness studies 1 , 168181 .
freeman , w.j . , 2003 ,
a neurobiological theory of meaning in perception , part i : information and meaning in nonconvergent and nonlocal brain dynamics .
international journal of bifurcation and chaos 13 , 24932511 . | we present results from numerical studies of supervised learning operations in recurrent networks considered as graphs , leading from a given set of input conditions to predetermined outputs .
graphs that have optimized their output for particular inputs with respect to predetermined outputs are asymptotically stable and can be characterized by attractors which form a representation space for an associative multiplicative structure of input operations . as the mapping from a series of inputs onto a series of such attractors generally depends on the sequence of inputs ,
this structure is generally non - commutative .
moreover , the size of the set of attractors , indicating the complexity of learning , is found to behave non - monotonically as learning proceeds . a tentative relation between this complexity and the notion of pragmatic information
is indicated . |
combining published data with high - quality vlt / fors spectroscopy of sample of fornax s0s ( bedregal et al .
2006a ) we have carried out a combined study of the tully - fisher relation and the stellar populations of these galaxies . despite the relatively small sample and the considerable technical challenges involved in determining the true rotation velocity @xmath1 from absorption line spectra of galaxies with significant non - rotational support ( see mathieu et al .
2002 ) , some very interesting results arise .
s0s lie systematically below the spiral galaxy tully - fisher relation in both the optical and near - infrared ( figure 1 ) . if s0s are the descendants of spiral galaxies , this offset can be naturally interpreted as arising from the luminosity evolution of spiral galaxies that have faded since ceasing star formation .
moreover , the amount of fading implied by the offset of individual s0s from the spiral relation seems to correlate with the luminosity - weighted age of their stellar population , particularly at their centres ( figure 2 ) .
this correlation suggests a scenario in which the star formation clock stopped when gas was stripped out from a spiral galaxy and it began to fade into an s0 .
the stronger correlation at small radii indicates a final last - gasp burst of star formation in this region .
see bedregal , aragn - salamanca & merrifield ( 2006b ) for details .
-band tully - fisher relation ( tfr ) for s0 galaxies using different samples from the literature ( open symbols ) and our vlt fornax data ( filled circles ) .
the solid and dashed lines show two independent determinations of the tfr relation for local spirals . on average (
dotted line ) , s0s are @xmath2 times fainter than spirals at similar rotation velocities ( bedregal , aragn - salamanca & merrifield 2006b ) . , width=384,height=316 ]
-band spiral tfr versus the stellar population age at the galaxy centre ( left panel ) , at @xmath3 ( middle panel ) and at @xmath4 ( right panel ) .
the lines show models for fading spirals .
note that the correlation is strongest for the central stellar populations of the galaxies , suggesting that the last episode of star formation took place there ( bedregal , aragn - salamanca & merrifield 2006b ) .
, height=158 ]
entirely consistent and independent evidence comes from our recent studies of the properties of the globular cluster ( gc ) systems and stellar populations of sos ( arag - salamanca , bedregal & merrifield 2006 ; barr et al .
if interactions with the intra - cluster medium are responsible for the transformation of spirals into s0s , the number of globular clusters in these galaxies will not be affected .
that is probably not true if more violent mechanisms such as galaxy - galaxy interactions are the culprit ( see , e.g. , ashman & zepf 1998 ) .
if we assume that the number of globular clusters remains constant , the gc specific frequency ( @xmath5number of gcs per unit @xmath6-band luminosity ) would increase due to the fading of the galaxy . on average ,
the gc specific frequency is a factor @xmath7 larger for s0s than it is for spirals ( aragn - salamanca et al .
2006 ) , meaning that in the process s0s become , on average , @xmath7 times fainter than their parent spiral .
furthermore , in this scenario the amount of fading ( or increase in gc specific frequency ) should grow with the time elapsed since the star formation ceased , i.e. , with the luminosity - weighted age of the s0 stellar population .
figure 3 shows that this is indeed the case , adding considerable weight to the conclusions reached from our tully - fisher studies .
in bedregal et al . ( 2007 ) we show that the central absorption - line indices in s0 galaxies correlate well with the central velocity dispersions in accordance with what previous studies found for elliptical galaxies . however , when these line indices are converted into stellar population properties , we find that the observed correlations seem to be driven by systematic age and alpha - element abundance variations , and not changes in overall metallicity as is usually assumed for ellipticals .
these correlations become even tighter when the maximum circular velocity is used instead of the central velocity dispersion .
this improvement in correlations is interesting because the maximum rotation velocity is a better proxy for the s0 s dynamical mass than its central velocity dispersion .
finally , the @xmath8-element over - abundance seems to be correlated with dynamical mass , while the absorption - line - derived ages also correlate with these over - abundances .
these correlations imply that the most massive s0s have the shortest star - formation timescales and the oldest stellar populations , suggesting that mass plays a large role in dictating the life histories of s0s .
the stellar populations , dynamics and globular clusters of s0s provide evidence consistent with these galaxies being the descendants of fading spirals whose star formation ceased .
however , caution is needed since significant problems could still exist with this picture ( see , e.g. , christlein & zabludoff 2004 ; boselli & gavazzi 2006 ) .
moreover , the number of galaxies studied here is still small , and it would be highly desirable to extend this kind of studies to much larger samples covering a broad range of galaxy masses and environments . of the luminosity - weighted ages is gyr vs. the globular cluster specific frequency ( @xmath9 ) of s0s .
the line shows the evolution expected for a fading galaxy according to the stellar population models of bruzual & charlot ( 2003 ) .
the correlation between the fading of the galaxies ( or increase in @xmath9 ) and the spectroscopically - determined age of their stellar populations is clearly consistent with the predictions of a simple fading model .
note that the @xmath9 value for ngc3115b is very unreliable and almost certainly severely overestimated due to contamination from the gc systems of neighbouring galaxies .
see barr et al .
( 2007 ) for details .
, height=278 ] | the stellar populations in the bulges of s0s , together with the galaxies dynamics , masses and globular clusters , contain very interesting clues about their formation .
i present here recent evidence suggesting that s0s are the descendants of fading spirals whose star formation ceased .
lenticular , or s0 , galaxies make up some 25% of large galaxies in the local universe ( dressler 1980 ) , so understanding how they form must constitute a significant element of any explanation of galaxy evolution .
their location at the crossroads between ellipticals and spirals in hubble s tuning - fork diagram underlines their importance in attempts to develop a unified understanding of galaxy evolution , but also means that it is not even clear to which of these classes of galaxy they are more closely related .
one often - cited piece of evidence comes from the fact that the proportion of s0s is substantially smaller in distant ( @xmath0 ) clusters than in nearby ones , while spirals show the opposite trend ( dressler et al .
1997 ) , strongly suggesting a transformation from one to the other . however , even if this scenario is accepted , it does not answer the question as to whether s0s are more closely related to spirals or ellipticals , which is intimately connected to the mechanism of transformation . if the transformation simply involves a spiral galaxy losing its gas content through ram pressure stripping ( gunn & gott 1972 ) or `` strangulation '' ( larson et al .
1980 ) , so ceasing star formation and fading into an s0 , then clearly s0s and spirals are closely related .
however , it is also possible that mergers can cause such a transformation : while equal - mass mergers between spirals create elliptical galaxies , more minor mergers can heat the original disk of a spiral and trigger a brief burst of star formation , using up the residual gas and leaving an s0 . in such a merger scenario , the mechanism for creating an s0 is much more closely related to that for the formation of ellipticals .
clues to which mechanism is responsible are to be found in the `` archaeological record '' that can be extracted from spectral observations of nearby s0s
. in particular , the present - day stellar dynamics should reflect the system s origins , with the gentle gas stripping of a spiral resulting in stellar dynamics very similar to the progenitor spiral , while the merger process will heat the stars , resulting in kinematics more dominated by random motions , akin to an elliptical .
in addition , the absorption line strengths can be interpreted through stellar population synthesis to learn about the metallicity and star formation histories of these systems .
even more interestingly , these dynamical and stellar properties can be compared to see if a consistent picture can be constructed for the formation of each system .
i present here some recent evidence suggesting that such a consistent picture is indeed emerging . |
streamer discharges determine the early stages of sparks and lightning ; they occur in various ionizable media in a large range of pressures and temperatures . in corona applications ,
they are widely used for surface charging in copiers and printers , for the removal of dust in electric precipitators and for ozone production and gas and water cleaning @xcite .
streamers underlie the operation of spark plugs in internal combustion engines , and there are new initiatives to improve the control of ignition @xcite .
another new application is flow control in aviation @xcite .
streamers also play a role in the ignition of energy efficient high pressure metal halid lamps @xcite . obviously , pressure , temperature and gas composition vary from one phenomenon to the next . in electrical engineering , the higher densities of liquids are desirable for fast switching , but pre - existing microbubbles largely influence streamer properties in fluids @xcite . for this reason ,
the operation of spark gaps filled with supercritical fluids is now under investigation @xcite .
we also recall experimental work on streamers in fluid nitrogen and argon @xcite .
streamer - like phenomena also occur in fast semiconductor switches @xcite . while in all these cases , the density of neutral particles is of the order of or considerably larger than in air at standard temperature and pressure , much lower densities are of interest in geophysics . at altitudes of 40 - 90 km in the atmosphere ,
enormous sprite discharges @xcite were found to evolve above thunderclouds .
they are thought to be related to streamer discharges through similarity laws @xcite as elaborated in the next section
. bar . number and diameters of streamers are similar in the lower panels at some distance from the needle electrode .
the pictures show the complete streamer evolution , hence the streamers in panel ( a ) do not reach the plate electrode .
note also , that in the extreme zoom of panel ( d ) , the pixel structure of the camera becomes visible and diameter evaluations become inaccurate .
the diameters reported in this paper are evaluated from pictures where the camera is moved physically so close to the discharge that such artifacts are avoided @xcite.,width=604 ] similarity laws between streamer discharges at different densities @xmath6 of the neutral gas or other medium follow from the fact that the basic length scale of the streamer discharge is the mean free path @xmath7 of the electron which is inversely proportional to the density of the medium : @xmath8 .
an electron is accelerated over this length by the local electric field * e * before the next collision with a neutral particle ; on this length it gains the energy @xmath9 where e is the elementary charge ; this energy has to be compared to the ionization energy of the neutral particle on which the electron impacts .
similarity for varying density @xmath6 therefore implies that the reduced electric field @xmath10 is the same at similar places , that the streamer velocity @xmath11 as well as velocity and energy distributions of individual electrons are the same at similar places , that all reduced lengths , i.e. , the ratio of lengths @xmath12 over the mean free electron paths @xmath13 are the same , and that the same electric potential is applied over the same reduced lengths these similarities have motivated the introduction of the unit townsend for the ratio @xmath10 of electric field over density , and we therefore will denote these similarity laws as townsend scaling .
the similarity can be tested on the experimental results as demonstrated in fig .
[ drukzoom ] , for a detailed description of these experiments , we refer to later chapters .
a minimal model @xcite for negative streamers with drift and diffusion of electrons , local impact ionization and space charge effects obeys these similarity laws perfectly .
the same is true when the model is extended by nonlocal photo - ionization in the low density limit ; this photo - ionization is commonly thought to explain the propagation of positive streamers in air @xcite .
corrections to the similarity laws come from three different sources .
+ @xmath14 the similarity laws are based on the fast two - particle processes between a charged particle and one of the abundant neutral particles that dominate the dynamics in the streamer ionization front . at higher densities , three - body processes inside the gas volume
can become important .
important examples include collisional quenching of the excited states that otherwise lead to photo - ionization in nitrogen - oxygen mixtures like air , as well as electron attachment to oxygen or electron ion recombination @xcite ; in all these cases the conservation of both momentum and energy requires the presence of a third particle , and the similarity laws break down on spatial or temporal scales where these processes become important .
+ @xmath15 corrections also come from the presence of electrodes or other matter ( like dust particles or gas bubbles ) whose lengths do not change when the gas density changes @xcite .
+ @xmath16 finally , the total number of charged particles in similar streamers varies like @xmath17 @xcite , therefore similar streamers at higher densities @xmath6 contain a lower total number of charged particles , and discrete particle fluctuations are more pronounced .
+ due to @xmath14 and @xmath15 , similarity laws for positive streamers in air are expected to hold only well below 30 torr , i.e. , 40 mbar ( 1 bar equals @xmath4 pa ) , where collisional quenching of photo - ionizing nitrogen states is negligible @xcite , in the streamer head when it is far enough from the electrodes , and at time scales when the loss of streamer conductivity due to electron attachment is still negligible .
we remark that the predicted propagation of positive streamers in absolutely pure nitrogen depends on the model input about photo - ionization in this gas , but the purity of our nitrogen is only 99.9% , therefore the common photo - ionization model for nitrogen - oxygen mixtures is still likely to apply .
the basic physical similarity variable is the density @xmath6 of neutral particles , therefore reduced fields and lengths should be defined as @xmath10 and @xmath18 .
the density @xmath6 is related to the pressure @xmath0 approximately as @xmath19 according to the ideal gas law , where @xmath20 is the boltzmann constant and @xmath21 is temperature . as our experiments are all performed at room temperature and as we measure pressure directly , we discuss the pressure dependence of our laboratory experiments rather than their density dependence and use @xmath22 and @xmath23 .
we extrapolate sprite measurements to room temperature using the ideal gas law .
experimental investigations of the density dependence of streamers are rare .
next to own preliminary visualizing work @xcite , pancheshnyi _ et al . _
@xcite have varied the pressure in the range of 0.5 to 1 bar in a 2 cm point - plane gap in synthetic air at a fixed voltage of 24 kv ; they find that the streamer diameter increases from 0.4 mm to 2.6 mm with decreasing pressure .
et al . _
@xcite have presented yet unpublished results in the pressure range of 0.01 to 0.06 bar at a recent workshop @xcite .
streamers in air in a dc corona at pressures of 1 - 10 bar were observed by marode _
et al . _
they conclude that the general discharge structures and , in particular , the streamer velocities , are similar when pressure changes . in sprite discharges , channels have been imaged with a telescope @xcite .
channels of varying diameter at fixed height are reported , e.g. , 60 to 145 m ( @xmath24 12 m ) in the altitude range of 60 - 64 km ( @xmath24 4.5 km ) and 20 - 50 m ( @xmath24 13 m ) in the altitude range of 76 - 80 km ( @xmath24 5 km ) @xcite .
the variation of sprite diameters resembles the variation of streamer diameters at 1 bar @xcite .
the sprites extend downward with velocities that vary by more than two orders of magnitude , ranging from 0.1 mm / ns (= @xmath4 m / s ) to more than 3@xmath25 m / s @xcite or even @xmath26 m / s @xcite .
other experimental work is concerned with the influence of voltage and gas composition at standard temperature and pressure on positive streamers . in previous and forth coming work @xcite
, we have shown that diameter and velocity of positive streamers in air at standard temperature and pressure in point - plane gaps of 4 and 8 cm can vary by more than an order of magnitude as a function of voltage , i.e. , from 0.2 to 3 mm and from 0.1 to 4 mm / ns for voltages between 5 and 96 kv ; furthermore , the diameters and velocities depend not only on the maximal voltage , but also on voltage rise time and possible current limitations @xcite
. a range of diameters is also presented by ono _
et al . _
@xcite where streamer diameters increase from 0.4 mm to 1.2 mm in a 1.3 cm point - plane gap in synthetic air at 1 bar when the voltage increases from 13 kv to 27 kv .
the average streamer velocity is 0.5 mm / ns at 18 kv . in nitrogen
they find the diameter to increase from 0.2 mm to 0.4 mm and the velocity to increase from 0.2 mm / ns to 0.5 mm / ns when the voltage increases from 16 kv to 27 kv and from 18 kv to 27 kv , respectively .
streamers in a protrusion - plane gap of 13 cm at 1 bar for a range of nitrogen - oxygen ratios and in the voltage range of 83 to 123 kv are described by yi and williams @xcite .
they observe an increase in propagation velocity from about 0.3 mm / ns to 4 mm / ns for increasing oxygen concentration ( up to 10% ) and increasing applied voltage .
obviously , there is a need to classify these phenomena . within the present paper
, we will investigate experimentally whether there are similarity laws between positive ( cathode directed ) streamers in the same gas type at different densities .
we will explore the density regime corresponding to pressures of 0.013 to 1 bar at room temperature and we will discuss whether our results over nearly two decades of density extrapolate to sprite discharges at much lower densities . if streamers at different densities are similar , then the huge sprite discharges at very low pressures can be investigated in small laboratory experiments at much higher pressure , and the fast temporal evolution of streamers in turn can be conveniently observed in slow motion and under the magnifying glass at lower pressures ; for the extreme case of sprites this has been recently demonstrated in @xcite .
most of our measurements have been performed in ambient air as it is most common in applications and in nature . however , air is a compound gas with 78% n@xmath27 , 20% o@xmath27 , 1% h@xmath27o , 0.03% co@xmath27 and 1% noble gases ; processes like photo - ionization between nitrogen and oxygen and electron attachment to oxygen have to be taken into account . for this reason
also experiments in nitrogen as a simple gas are performed and presented ; however , we here only present results at a nitrogen purity of about 99.9 % , where impurities still seem to play a role .
this paper concentrates on measurements of streamer diameters , and also reports the size of the initial ionization cloud at inception and the velocities , and it presents first results on branching lengths .
branching angles were recently measured in @xcite . as shown in @xcite and discussed in more detail in @xcite
, there is a critical inception voltage required for streamer formation ; at the inception voltage , streamers at a given pressure always appear to attain the same minimal diameter .
when a higher voltage is applied within a sufficiently short rise time , considerably thicker and faster streamers are formed @xcite . however , when the voltage is increased sufficiently slowly , streamers emerge as soon as the inception voltage is reached , and they ( obviously ) have minimal diameter . in the present paper , we have chosen to test the similarity laws on these thinnest streamers for several reasons .
first , it appears that there is a lower limit for the streamer diameter but an upper limit is not established @xcite .
therefore , the minimal diameters at each pressure are appropriate quantities to compare .
second , the thinnest streamers are furthest away from the electrodes that might break the similarity laws locally as discussed in section [ townsend ] .
third , the minimal streamers do not evolve into glow and/or spark as easily as the thick ones , and therefore they can be imaged more easily .
the paper is organized as follows .
section 2 contains a detailed description of the experimental setup , section 3 presents the measurements of streamer diameters and of the size of the initial cloud , of streamer velocities and of branching ratios .
section 4 contains discussion , conclusions and outlook .
appendix a illustrates the difference between streamers in air and in our n@xmath27-o@xmath27 mixture .
= charging resistor , @xmath28 = rise time determining resistor , @xmath29 = decay time determining resistor , t.o.= to oscilloscope.,width=604 ] .values of the components used in the c - supply . for @xmath30 , @xmath28 , @xmath29 , shunt and cc , see figure [ expsetupoverview ] , @xmath31 = voltage rise time , sc = behlke hts 651 semiconductor switch ; sg = homemade sparkgap . [ cols="<,<,<,<,<,<,<,<",options="header " , ] the dimensions of the cloud in air increase with decreasing pressure or increasing voltage .
the cloud is generally oblate with an aspect ratio of @xmath32 0.7 . at 0.1 and 0.2
bar , a different aspect ratio of @xmath32 1.6 is seen at the lowest voltages .
the height of the cloud at the inception voltage roughly scales with inverse pressure ; this leads to a reduced value of @xmath33 mm@xmath34bar . in n@xmath27 ,
the cloud is smaller and less regular than in air , as can , e.g. , be seen in figure [ n2airsimpel ] . only at the lowest pressures of 0.015 bar in a 16 cm gap ,
the sphere is clearly observed with a size similar to the one in air , cf .
[ gevoeltijdsopg ] . at 1 bar at a voltage of 49 kv ( @xmath35 mm@xmath34bar ) .
the gate delay @xmath21 and exposure time @xmath36 are indicated in each panel .
the streamers reach the plate electrode some time after the exposure of panel ( c ) .
when the same experiment is performed in air at 1 bar , the streamers do not cross the gap.,title="fig:",width=188 ] ( a ) @xmath37s , @xmath38s at 1 bar at a voltage of 49 kv ( @xmath35 mm@xmath34bar ) .
the gate delay @xmath21 and exposure time @xmath36 are indicated in each panel .
the streamers reach the plate electrode some time after the exposure of panel ( c ) .
when the same experiment is performed in air at 1 bar , the streamers do not cross the gap.,title="fig:",width=188 ] ( b ) @xmath39s , @xmath40s at 1 bar at a voltage of 49 kv ( @xmath35 mm@xmath34bar ) .
the gate delay @xmath21 and exposure time @xmath36 are indicated in each panel .
the streamers reach the plate electrode some time after the exposure of panel ( c ) .
when the same experiment is performed in air at 1 bar , the streamers do not cross the gap.,title="fig:",width=188 ] ( c ) @xmath41s , @xmath40s streamer velocities are evaluated from single photographs like panel ( a ) of figure [ figdiag ] .
the method is based on the observation that for exposure times as short as 1 ns , the actively growing streamer head in air and nitrogen is visible as a glowing ball @xcite , this feature recently also has been demonstrated for sprites @xcite . extended lines therefore indicate the trace of the propagating streamer head within a longer exposure time .
delay @xmath21 and exposure time @xmath36 of the photographs are chosen in such a way that velocities conveniently can be determined , see below .
we evaluate the velocities here only in the 16 cm gap and at three different locations ; near the anode tip , halfway the gap and near the cathode plate .
the velocities are determined from the longest streamers in each picture since they propagate most parallel to the camera s focal plane .
( this constraint could be dropped in future measurements when using stereographic imaging @xcite . )
note that the velocities are related to the diameters @xcite .
however , the diameters can not be accurately determined from the full photographs of the 16 cm gap as the camera resolution is insufficient , cf . the extensive discussion of imaging artifacts in @xcite that is briefly recalled in section [ chn2airdiam ] .
typical exposure times @xmath36 for the velocity measurements are 50 to 500 ns .
these values are chosen in such a way that the propagation distance is short enough to contain little or no branching , but also long enough to limit relative errors in d@xmath42 and @xmath36 due to the extension of the streamer head , due to the emission time of the excited c - state of nitrogen ( @xmath432 ns ) that emits the light , and due to inaccuracies of the camera s gate time @xmath36 .
therefore the exposure time is adjusted such that the streamer trace d@xmath42 is about 1 to 4 cm in the 16 cm gap as in fig .
[ vel ] . at 0.015 mbar
the development of the streamer is so limited that its velocity can not be obtained , cf .
figure [ gevoeltijdsopg ] ; a much larger vessel would be required for this purpose . the comparison of panels ( a ) and ( b ) in figure [ vel ] shows , that the propagation length is roughly the same while the exposure time is five times longer .
this demonstrates the general feature that the velocity decreases when the streamer propagates down the gap into regions with lower background field ; in fig . [ vel ] the velocity change between different gap locations is in fact larger than in other cases .
[ n2airvel ] presents the range of streamer velocities as a function of pressure and voltage .
the bars in the figure extend from the lowest velocity ( typically measured near the cathode plate ) to the highest velocity ( typically near the anode tip ) .
( 1 mm / ns = @xmath44 m / s ) .
the bars indicate the velocity range observed at different locations in the gap.,title="fig:",height=226 ] ( a ) ( 1 mm / ns = @xmath44 m / s ) .
the bars indicate the velocity range observed at different locations in the gap.,title="fig:",height=226 ] ( b ) the jitter in the discharge inception makes it here difficult as well to obtain images that can be analyzed with confidence .
therefore the number of time resolved photographs per parameter set is limited to only a few .
furthermore , at low voltages , one must be sure that the streamer on the image finishes because the camera closes , and not because the streamer stops anyway .
in air this constraint is particularly severe , as the maximal streamer length is typically shorter than in nitrogen , as is shown in figure [ n2airvergelijk ] .
for all these reasons , it is uncertain whether the velocity of streamers of minimal diameter is contained in our measured data set . in air at 1 bar
the minimal velocity is of the order of 0.1 mm / ns = @xmath4 m / s ; far from the needle electrode it even decreases to @xmath45 m / s .
a value of @xmath4 m / s is also found as the minimal velocity in a 40 mm gap @xcite . at 0.4 bar
the smallest velocity takes the higher value of @xmath46 m / s , but then the diameter is clearly not minimal , but has the value of @xmath47 mm@xmath34bar , cf . fig .
6.9 a in @xcite . in our n@xmath27
the situation is more complicated .
the velocities at 1 bar are much lower than in air , e.g. , from figure [ vel ] ( c ) , one can easily extract a velocity of @xmath48 m / s .
but at lower pressure the velocities in nitrogen are the same or even higher than in air ; here a pressure dependent impurity level might play a role .
streamer branching can be characterized by the branching angle and the streamer length between branching . for the determination of the branching angle through stereographic imaging
, we refer to @xcite . here
we measure the length @xmath49 of the streamer from one branching event to the next .
we relate this length to the streamer diameter @xmath50 in the branching ratio @xmath51 , both quantities are illustrated in panel ( b ) of figure [ figdiag ] . for this evaluation
, photographs must show the complete discharge as well as accurate diameters without instrumental broadening , therefore only data from a 4 cm gap are evaluated .
the diameter is measured just before the branching of the channel ; the streamer length between branching events is measured as the starting point of one streamer branch to the starting point of the next one .
evaluations in the region close to the electrode tip are mostly avoided since the tip region is frequently overexposed and therefore the actual starting point of the streamer can not be determined accurately . for the interpretation of our data
it should be noted that streamers do not contribute to our statistics if they do not branch ; this is the case , e.g. , at low pressure as in fig .
[ drukzoom ] ( c ) or far from the electrodes as in fig .
[ vel ] ( c ) . in our n@xmath27 , the branches are often very short , cf . fig .
[ n2airsimpel ] ( b ) . as a blurry surrounding , scattered light or accidentally overexposed pixels could be misinterpreted as a short branch , only streamers with sharp , bright branches are evaluated in our n@xmath27 which limits our data set . of branching lengths @xmath49 over streamer diameters @xmath50 at different pressures in a 4 cm gap just above the respective inception voltage .
the points indicate the mean value , the bars indicate the range of the statistical distribution ; it must not be regarded as an error bar .
b ) the branching length @xmath49 as a function of the streamer diameter @xmath50 in an 8 cm gap in air at 1 bar and different voltages . here
data from @xcite with fast voltage rise time and thick streamers are plotted.,title="fig:",height=226 ] ( a ) of branching lengths @xmath49 over streamer diameters @xmath50 at different pressures in a 4 cm gap just above the respective inception voltage .
the points indicate the mean value , the bars indicate the range of the statistical distribution ; it must not be regarded as an error bar .
b ) the branching length @xmath49 as a function of the streamer diameter @xmath50 in an 8 cm gap in air at 1 bar and different voltages . here data from @xcite with fast voltage rise time and thick streamers are plotted.,title="fig:",height=226 ] ( b ) figure [ n2airbranch ] ( a ) shows the branching ratio @xmath51 for several pressures and for voltages just above the inception voltage .
the streamer diameters in this case are small but not completely minimal .
the symbols in the figure ( black dots for air and red squares for our n@xmath27 ) indicate the median of the data .
the bars indicate the range from the lowest to the highest value and therefore the statistical distribution of the branching lengths @xmath49 . that there is actually a statistical distribution can be seen , e.g. , in figure [ figdiag ] ( b ) where the channel indicated by @xmath49 branches only once while the streamer on the right branches two times over a shorter distance .
if streamers at different pressures and otherwise similar parameters are similar , then the length ratio @xmath51 does not depend on pressure . within the scatter of the data ,
this is indeed the case ; a horizontal line @xmath52 fits the data for air , and @xmath53 fits the data for our n@xmath27 . here
the error is calculated as the standard deviation of the points in the figure .
figure [ n2airbranch ] ( b ) presents data for thicker streamers in air at 1 bar . here the experimental results from @xcite are evaluated . as in these experiments
voltages of up to 60 kv were applied , and as the voltage rise time was as short as 25 ns , streamers with diameters in the range from 0.2 to 3.25 mm were created .
the figure shows the branching length @xmath49 as a function of the streamer diameter @xmath50 .
the branching length increases with the streamer diameter ; it can be fitted with the line @xmath54 . for comparison ,
the streamers of minimal diameter in air at 1 bar are formed at a voltage of about 10 kv ; their branching ratio has the larger value @xmath55 as shown in panel ( a ) of the figure .
we conclude that at least in air at 1 bar the absolute length @xmath49 between branches increases with diameter @xmath50 and voltage @xmath56 , but that the ratio @xmath51 of branching length over streamer diameter slightly decreases with diameter and voltage .
the primary aim of this paper is to test the similarity laws on streamers at different pressures @xmath0 , or more precisely , for streamers at different densities @xmath6 , where @xmath19 according to the ideal gas law . for the streamer diameter
, we have to analyze its reduced value @xmath57 . however , even at fixed pressure
, the streamer diameter can vary by more than an order of magnitude , depending on applied voltage and voltage rise time @xcite .
therefore we concentrate on the minimal diameter @xmath1 . for the reduced minimal streamer diameter in air
, we find in our experiments the constant value @xmath58 at room temperature when pressure changes from 0.013 to 1 bar .
this value agrees with old measurements of repetitive positive streamers in ambient air with dc voltage yielding diameters of the order of @xmath59 for the full width at half maximum ( fwhm ) @xcite ; the dc drive is an extreme case of slow voltage rise ; consistently with our findings , it creates streamers of minimal diameter .
it is quite unexpected that the reduced streamer diameters in air in this pressure range turn out to be constant . as discussed in section [ townsend ] , deviations from the similarity law
are expected above @xmath60 mbar , because the excited nitrogen states that are responsible for photo - ionization , increasingly loose their energy through collisions with neutral molecules rather than through photo - emission .
it is also remarkable that the smallest reduced diameter of a sprite channel obtained in the telescopic images in @xcite is @xmath61 .
it is consistent with our laboratory value even though the pressure was as low as @xmath62 bar , cf .
[ n2aireddie ] . of course , it is not clear whether sprites of minimal diameter were included in the limited data set of @xcite ; and the higher background ionization at sprite altitude might lead to corrections to the similarity laws . in the recent literature ,
the diameters of positive streamers in air are all within the range from 0.2 to 3 mm@xmath34bar that was discussed in @xcite , but no streamers of minimal diameter were reported elsewhere .
et al . _
@xcite report measurements and simulations with @xmath63 mm@xmath34bar at 1 bar and @xmath64 kv/(cm@xmath34bar ) , and with @xmath66 mm@xmath34bar at 0.5 bar and @xmath67 kv/(cm@xmath34bar ) .
a stagnating streamer had a diameter of 0.4 mm at 1 bar in @xcite while the same authors find streamer diameters between 1 and 3 mm at 1 bar in stronger fields in @xcite , both in experiments and in simulations .
liu and pasko @xcite report simulations with @xmath68 = 1.0 , 1.4 and 1.2 mm@xmath34bar at 1 , 0.014 and @xmath69 bar and @xmath70 kv/(cm@xmath34bar ) . for our n@xmath27 with an impurity level of about 0.1 %
, we find a reduced minimal diameter of @xmath71 at room temperature , but the spread of the data is larger than in air .
this is probably due to temporal variations of the impurity concentration ; a new experimental setup is therefore now in development .
the similarity laws state that the velocities of streamers with the same reduced diameter are the same .
we concentrate on streamers with minimal diameter .
unfortunately , their velocity is very difficult to determine at pressures below 1 bar .
this is due to the increasing jitter of the inception time which makes appropriate timing of the iccd photography increasingly difficult .
a second complication is that it is difficult to ascertain that the streamer segment ends because the exposure time of the camera finishes , and not because the streamer stops by itself .
the minimal velocity at 1 bar in air is 0.1 mm / ns = @xmath4 m / s , the higher values at lower pressure in figure [ n2airvel ] do not belong to streamers of minimal diameter .
it is remarkable that the lowest sprite velocity reported in @xcite is @xmath4 m / s as well .
however , this lower limit is reported only occasionally for sprites while values of up to @xmath26 m / s are observed @xcite . now with fast voltage rise time , streamer diameters and velocities also increase to 3 mm and @xmath72 m / s for a voltage of 96 kv in ambient air @xcite ; probably they increase further with higher applied voltage
. therefore maximal velocities of streamers or sprites can not be compared while minimal velocities are well defined and similar .
the streamer velocities in our n@xmath27 are quite different . at 1 bar
they are much lower , down to @xmath73
m / s , but at 0.4 and 0.065 bar the values are quite the same or even higher than in air .
again we might not observe minimal streamers at the lower pressures , and the influence of the impurities again is uncertain and might increase with decreasing pressure . as positive streamers can only propagate due to photo - ionization , due to electron detachment from o@xmath74 or due to background ionization , and as the photo - ionization rate decreases and the photo - ionization lengths increase with decreasing oxygen concentration @xcite , quite different velocities can be expected in our n@xmath27 .
whether positive streamers in pure nitrogen without background ionization propagate at all , is not known .
the only other data is presented in @xcite , it is not conclusive .
the discharge starts at the anode point as an oblate cloud with an aspect ratio of height over width @xmath75 .
this cloud is clearly observed in air at pressures of 0.4 bar and lower . at 1 bar
the cloud is difficult to distinguish as it is very small , and the iccd images are often overexposed at the tip .
the cloud size at the inception voltage approximately obeys a similarity law ; the reduced cloud height is approximately @xmath76 . in our n@xmath27 ,
the cloud is less pronounced .
elsewhere we have measured the distribution of branching angles @xcite .
here we have introduced and measured the branching ratio @xmath51 to characterize the shape of the streamer tree ; here @xmath49 is the distance between branching events and @xmath50 the streamer diameter just before branching . in air
, this ratio is @xmath77 for streamers of minimal diameter for all pressures in the range from 0.065 to 1 bar . therefore the overall branching pattern is quite independent of pressure , even though the stochastic fluctuations are larger at higher pressures as discussed in section [ townsend ] . if random ionization avalanches would play a major role in the branching process ( cf .
the discussion in section 5 of @xcite ) , we would expect relatively more branching , i.e. , a smaller @xmath51 , at higher pressures . for thick streamers in air
at 1 bar , we find @xmath51 = 8 @xmath24 4 .
therefore for fixed pressure , thicker streamers are longer from one splitting to the next , but measured on the scale of their diameter , they branch slightly more frequently .
this trend agrees with observations of yi and williams @xcite ; they report an average branching distance of 3 mm for positive streamers with a diameter of about 2 mm in n@xmath27 at 1 bar at 98 kv .
this yields @xmath78 in qualitative agreement with our observation that the ratio @xmath51 decreases with increasing voltage at a fixed pressure .
it also indicates that streamers in nitrogen branch more frequently than in air .
the branching ratio in our n@xmath27 is @xmath79 for thin streamers in the same pressure range .
this value is somewhat lower than in air . as also the diameter @xmath50 of n@xmath27 streamers
is smaller , the absolute streamer length @xmath49 between branching is only half of that in air at the same pressure .
this can be seen qualitatively in figure [ n2airsimpel ] . for a further comparison of streamers in air and in our n@xmath27
, we refer to the [ app ] .
we conclude that the reduced diameter @xmath57 of minimal streamers is constant for pressures in the range from @xmath62 to 1 bar at room temperature , while present theory predicts a change of behavior above approximately 0.04 bar .
the velocity , initial cloud size and branching structure also scale quite well with pressure , but due to the experimental limitations this could only be tested in a less extended pressure range .
evaluating the results in our n@xmath27 is somewhat problematic because presently the impurity level in the gas is not sufficiently controlled . in the near future we plan new experiments to improve this situation . in any case , the comparison of our experimental results with theory at varying pressure and gas composition and the discussion of the microscopic physical mechanisms will pose challenges for the future .
this work was supported by stw under contract number 06501 and by nwo under contract number 047.016.017 .
( right column ) at @xmath80 bar and @xmath81 kv ( @xmath82 mm@xmath34bar , upper row ) and at @xmath83 bar and @xmath84 kv ( @xmath85 mm@xmath34bar , lower row ) .
the camera delay is @xmath86 , the exposure time @xmath36 and pulse frequency are : a ) 0.6 @xmath87s , single shot ; b ) 1.5 @xmath87s , 1 hz ; c ) 0.7 @xmath87s , 1 hz ; and d ) 0.53 @xmath87s , 10 hz .
a part of the streamers eventually reaches the plate electrode , but only after the exposure times of photos ( a ) , ( c ) and ( d).,title="fig:",height=226 ] a ) air , 1 bar , 21 kv + + ( right column ) at @xmath80 bar and @xmath81 kv ( @xmath82 mm@xmath34bar , upper row ) and at @xmath83 bar and @xmath84 kv ( @xmath85 mm@xmath34bar , lower row ) .
the camera delay is @xmath86 , the exposure time @xmath36 and pulse frequency are : a ) 0.6 @xmath87s , single shot ; b ) 1.5 @xmath87s , 1 hz ; c ) 0.7 @xmath87s , 1 hz ; and d ) 0.53 @xmath87s , 10 hz .
a part of the streamers eventually reaches the plate electrode , but only after the exposure times of photos ( a ) , ( c ) and ( d).,title="fig:",height=226 ] c ) air , 0.1 bar , 5 kv ( right column ) at @xmath80 bar and @xmath81 kv ( @xmath82 mm@xmath34bar , upper row ) and at @xmath83 bar and @xmath84 kv ( @xmath85 mm@xmath34bar , lower row ) .
the camera delay is @xmath86 , the exposure time @xmath36 and pulse frequency are : a ) 0.6 @xmath87s , single shot ; b ) 1.5 @xmath87s , 1 hz ; c ) 0.7 @xmath87s , 1 hz ; and d ) 0.53 @xmath87s , 10 hz .
a part of the streamers eventually reaches the plate electrode , but only after the exposure times of photos ( a ) , ( c ) and ( d).,title="fig:",height=226 ] b ) n@xmath27 , 1 bar , 19 kv + + ( right column ) at @xmath80 bar and @xmath81 kv ( @xmath82 mm@xmath34bar , upper row ) and at @xmath83 bar and @xmath84 kv ( @xmath85 mm@xmath34bar , lower row ) .
the camera delay is @xmath86 , the exposure time @xmath36 and pulse frequency are : a ) 0.6 @xmath87s , single shot ; b ) 1.5 @xmath87s , 1 hz ; c ) 0.7 @xmath87s , 1 hz ; and d ) 0.53 @xmath87s , 10 hz .
a part of the streamers eventually reaches the plate electrode , but only after the exposure times of photos ( a ) , ( c ) and ( d).,title="fig:",height=226 ] d ) n@xmath27 , 0.1 bar , 5 kv though nitrogen is the major component of air , streamers in nitrogen and air have two important distinctions on the microscopic level : @xmath88 nitrogen - oxygen mixtures have an efficient photo - ionization mechanism that allows streamers to propagate through a nonlocal ionization mechanism against the electron drift direction ( though the model parameters for this interaction are not very well known ) ; and @xmath89 free electrons are easily lost in air through attachment to oxygen .
therefore , it is quite instructive to compare discharges in these two gases .
note that a quantitative link between these microscopic mechanisms and the macroscopic appearance is emerging only now , except for stating that electron attachment limits the duration of discharges in air .
the differences between streamers in air and nitrogen are demonstrated on typical photographs taken in the gap of 40 and 160 mm .
another good example of long streamers in the 160 mm gap is given in @xcite .
we recall that our n@xmath27 contains about 0.1 % oxygen and our nitrogen - oxygen mixture 0.2 % oxygen ; the oxygen concentration in air is about 100 times higher than in our n@xmath27 or in our mixture .
figure [ n2airuiteinde ] shows photographs of streamers in air and in our n@xmath27 at pressures of 1 and 0.1 bar .
the comparison of panels ( a ) and ( b ) shows clearly that streamers in nitrogen move more in a zigzag manner while the streamers in air are straighter .
furthermore , the streamers in nitrogen branch more frequently .
the panels also show that streamers in nitrogen have a better contrast between in- and out - of - focus structures .
panels ( c ) and ( d ) show that the streamers in n@xmath27 are thinner , and that the light emitting cloud at the electrode tip is clearly visible in air while it is less pronounced in nitrogen .
the camera timing in panels ( c ) and ( d ) is such that the streamers are still propagating ; with a longer exposure time , one would see them reaching the cathode plate .
therefore , the different light emission structure of the propagating streamer tips can be studied on these panels : the streamer tips in air in panel ( c ) are round and stop abruptly , while the tips in n@xmath27 in panel ( d ) are more diffuse .
the length along the streamer over which the intensity drops to 50% of its maximal value is about 4 times longer in n@xmath27 .
bar and @xmath90 kv ( @xmath91 mm@xmath34bar ) ; ( a , b ) : air at 20 kv , and ( c , d ) : nitrogen - oxygen mixture ( 99.8 : 0.2 ) at 22 kv .
the oxygen concentration in air is about 100 times higher than in the mixture .
panels ( a ) and ( c ) are on the same intensity scale , ( b ) has a different scale to show the complete streamer structure .
panel ( d ) zooms into the region of 65 to 110 mm from the needle tip .
the figures show the full evolution , hence the streamers in air do not cross the gap.,title="fig:",height=226 ] ( a ) : air , oxygen concentration is 100 times higher than in the mixture ( c ) + + bar and @xmath90 kv ( @xmath91 mm@xmath34bar ) ; ( a , b ) : air at 20 kv , and ( c , d ) : nitrogen - oxygen mixture ( 99.8 : 0.2 ) at 22 kv .
the oxygen concentration in air is about 100 times higher than in the mixture .
panels ( a ) and ( c ) are on the same intensity scale , ( b ) has a different scale to show the complete streamer structure .
panel ( d ) zooms into the region of 65 to 110 mm from the needle tip .
the figures show the full evolution , hence the streamers in air do not cross the gap.,title="fig:",height=226 ] ( c ) : nitrogen - oxygen mixture ( 99.8:0.2 ) , + same view and intensity scale as in ( a ) bar and @xmath90 kv ( @xmath91 mm@xmath34bar ) ; ( a , b ) : air at 20 kv , and ( c , d ) : nitrogen - oxygen mixture ( 99.8 : 0.2 ) at 22 kv . the oxygen concentration in air is about 100 times higher than in the mixture .
panels ( a ) and ( c ) are on the same intensity scale , ( b ) has a different scale to show the complete streamer structure .
panel ( d ) zooms into the region of 65 to 110 mm from the needle tip .
the figures show the full evolution , hence the streamers in air do not cross the gap.,title="fig:",height=226 ] ( b ) : the data in air of ( a ) , + but with enhanced intensity scale + + bar and @xmath90 kv ( @xmath91 mm@xmath34bar ) ; ( a , b ) : air at 20 kv , and ( c , d ) : nitrogen - oxygen mixture ( 99.8 : 0.2 ) at 22 kv .
the oxygen concentration in air is about 100 times higher than in the mixture .
panels ( a ) and ( c ) are on the same intensity scale , ( b ) has a different scale to show the complete streamer structure .
panel ( d ) zooms into the region of 65 to 110 mm from the needle tip .
the figures show the full evolution , hence the streamers in air do not cross the gap.,title="fig:",height=226 ] ( d ) : same parameters as in ( c ) , + but camera has moved closer by .
figure [ n2airvergelijk ] shows streamers in air and in the nitrogen - oxygen mixture with ratio 99.8:0.2 with exactly the same setup and camera settings apart from the gas composition .
( the circular outer edge of the streamers visible in the figure is the edge of the quartz window in the vacuum vessel that limits the view when imaging the 16 cm gap , cf .
[ setup ] . )
the exposure time is 70 @xmath87s , which is longer than the voltage pulse duration , therefore the whole path of the streamers towards the cathode plate can be seen .
panels ( a ) and ( c ) show streamers in air and in the mixture with identical intensity coding ; the streamers in the mixture are much more intense , and they bridge the complete gap .
the rather faint data of panel ( a ) is reproduced with a more sensitive color coding in panel ( b ) ; here it can be verified that the streamers in air do not bridge the gap . the intensity difference between air and the mixture
is not related to the bridging of the gap , since it is also present at other pressures and voltages where the streamers in the mixture do not bridge the gap .
panels ( b ) and ( c ) show that the streamers in the mixture branch more .
note that the results in our n@xmath27 are similar .
panel ( d ) zooms into the region of 65 to 110 mm below the tip ( i.e. halfway the gap ) for the streamers in the mixture .
it shows many branches that mostly remain short .
we remark that we use figures such as panel ( d ) to measure diameters in gaps of 16 cm , since the resolution of panel ( c ) is insufficient and leads to measuring artifacts @xcite .
10 van veldhuizen e m 2000 _ electrical discharges for environmental purposes : fundamentals and applications _
( huntington , ny : nova science publishers ) , isbn 1 - 56072 - 743 - 8 , 420 pages yan k 2001 _ corona plasma generation _ , phd thesis , eindhoven university of technology , the netherlands .
available on www.tue.nl/bib pokryvailo a , wolf m , yankelevich y , wald s
, grabowski l r , van veldhuizen e m , rutgers w r , raiser m , glocker b , eckhardt t , kempenaers p and welleman a 2006 _ ieee trans .
plasma sci . _ * 34 * 1731 ebert u , montijn c , briels t m p , hundsdorfer w , meulenbroek b , rocco a and van veldhuizen e m 2006 s118 winands g j j , van heesch e j m , pemen a j m , liu z and yan k in @xcite and @xcite workshop _ streamers , sprites , leaders , lightning : from micro- to macroscales _ , lorentz center , leiden , the netherlands , oct .
8 - 12 , 2007 , abstracts and presentations available on http://www.lorentzcenter.nl/lc/web/2007/265/info.php3?wsid=265 cluster issue on `` streamers , sprites and lightning '' in , scheduled for aug .
2008 , guest editors : ebert u and sentman d d agneray a , auzas f , makarov m , jaffrezic x , puech v and tardiveau p 2007 _ xxviii icpig , july 15 - 20 , prague , czech republic _ paper 3p10 - 01 starikovskaia
s m 2006 r265 aleksandrov
n l , kirpichnikov a a , kindusheva s v , kosarev i n , starikovskii a y 2007 , _
45th aiaa aerospace sciences meeting , jan 2007 , nevada , usa _
paper aiaa-2007 - 0997 pitchford l 2007 _ xxviii icpig , july 15 - 20 , prague , czech republic _
paper g01 moss r s , eden j g and kushner m j 2004 2502 qian j , joshi r p , j. kolb j , schoenbach k h , dickens j , neuber a , butcher m , cevallos m , krompholz h , schamiloglu e and gaudet j 2005 _ j. appl .
phys . _ * 97 * 113304 an w , baumlung k and bluhm h 2007 _ j. appl .
phys . _ * 101 * 053302 lock e h , saveliev
a v and kennedy l a 2006 _ ieee trans . on plasma sci .
* 2467 dnat a 2006 _ ieee trans .
diel . and el . insul .
_ * 13 * 518 rodin p , ebert u , hundsdorfer w and grekhov i v 2002 _ j. appl .
_ * 92 * 1971 sentman d d , wescott e m , osborne d l and heavner m j 1995 _ geophys .
* 22 * 1205 pasko v p , stanley m a , mathews j d , inan u s and wood t g 2002 _ nature _ * 416 * 152 gerken e a , inan u s and barrington - leigh c p 2000 _ geophys .
* 27 * 2637 mcharg m g , stenbaek - nielsen h c , kammae t 2007 _ geophys . res .
lett . _ * 34 * l06804 stenbaek - nielsen h c , mcharg m g ,
kammae t and sentman d d 2007 _ geophys .
lett . _ * 34 * l11105 pasko v p 2007 _ plasma sources sci
. technol . _
* 16 * s13 briels t m p , kos j , van veldhuizen e m and ebert u 2006 5201 ebert u , van saarloos w and caroli c 1997 _ phys .
e _ * 55 * 1530 zheleznyak m b , mnatsakanyan a kh and sizykh s v 1982 _ high temp . _
* 20 * 357 morrow r and lowke j j 1997 614 kulikovsky a a 1997 1514 pasko v p , inan u s and bell t f 1998 _ geophys .
lett . _ * 25 * 2123 babaeva n yu , naidis g v 2001 j. electrostat . *
53 * 123 liu n and pasko v p 2004 _ j. geophys .
res . _ * 109 * a04301 luque a , ebert u , montijn c and hundsdorfer w 2007 _ appl .
lett . _ * 90 * 081501 achat s , teisseyre y and marode e 1992 _ j. phys .
phys . _ * 25 * 661 babaeva
n y , bhoj a n and kushner m j 2005 `` streamer dynamics in gases containing dust particles '' , _ xviith icpig , eindhoven , the netherlands _ file 04 - 400 , available on : www.icpig2005.nl
liu n and pasko v p 2006 _ j. phys .
d. : appl .
_ * 39 * 327 briels t m p , van veldhuizen e m and ebert u 2005 _ ieee trans .
plasma sci .
_ * 33 * 264 pancheshnyi s v , nudnova m m and starikovskii a yu 2005 _ phys .
e _ * 71 * 016407 becerra m , valente m , cooray v , williams e r and golka r `` laboratory experiments of streamers at air pressures ranging from 10 to 60 mbar '' , poster presented at @xcite briels t m p , winands g j j , van veldhuizen e m and ebert u 2008 `` positive and negative streamers in ambient air : measuring diameter , velocity and dissipated energy '' , submitted to @xcite , preprint : http://arxiv.org/abs/0805.1376 luque a , ratushnaya v and ebert u 2008 `` positive and negative streamers in ambient air : modeling evolution and velocities '' , submitted to @xcite , preprint : http://arxiv.org/abs/0804.3539 moudry d r , stenbaek - nielsen h c , sentman d d and wescott e m 2002 _ geophys .
lett _ * 29 * 1992 mcharg m g _ et al .
_ 2002 j. geophys .
107 * , issue a11 , 1364 ono r and oda t 2003 1952 yi w j and williams p f 2002 205 nijdam s , moerman j s , briels t m p , van veldhuizen e m , ebert u 2008 _ appl .
lett . _ * 92 * 101502 gravendeel b 1987 _ negative corona discharges : a fundamental study _ ,
phd thesis , eindhoven university of technology , the netherlands , available on www.tue.nl/bib wetzer j m and van der laan p c t 1989 _ ieee trans . electrical insulation _ * 24 * 297 van veldhuizen e m and rutgers w r 2003 _ j. phys .
d : appl . phys . _ * 36 * 2692 briels t m p , van veldhuizen e m and ebert u 2008 `` positive streamers in ambient air and a n@xmath27:o@xmath27-mixture ( 99.8 : 0.2 ) '' , _ ieee trans .
plasma sci .
_ [ to appear ] briels t m p , van veldhuizen e m and ebert u `` experiments on the diameter of positive streamers in air '' , _ xviith icpig , eindhoven , the netherlands _ , file 04 - 418 ( 2005 ) , available on : www.icpig2005.nl , http://www.auf.asn.au/meteorology/section1a.html#atmospheric structure http://msis.jsc.nasa.gov briels
t m p , van veldhuizen e m and ebert u 2008 `` time resolved measurements of streamer inception in air '' , _ ieee trans .
plasma sci .
_ [ to appear ] briels
t m p 2007 _ exploring streamer variability in experiments _ phd thesis at technische universiteit eindhoven gibert a , bastien f j 1989 1078 pancheshnyi s v , starikovskii a yu 2004 b1 pancheshnyi s v , starikovskii a yu 2003 2683 luque a , ebert u , hundsdorfer w 2008 `` interaction of streamers in air and other oxygen - nitrogen mixtures '' , revised for phys . rev .
, preprint : http://arxiv.org/abs/0712.2774 | positive streamers in ambient air at pressures from 0.013 to 1 bar are investigated experimentally .
the voltage applied to the anode needle ranges from 5 to 45 kv , the discharge gap from 1 to 16 cm . using a `` slow '' voltage rise time of 100 to 180 ns ,
the streamers are intentionally kept thin . for each pressure
@xmath0 , we find a minimal diameter @xmath1 . to test whether streamers at different pressures are similar , the minimal streamer diameter @xmath2 is multiplied by its pressure @xmath0 ; we find this product to be well approximated by @xmath3 over two decades of air pressure at room temperature .
the value also fits diameters of sprite discharges above thunderclouds at an altitude of 80 km when extrapolated to room temperature ( as air density rather than pressure determines the physical behavior ) .
the minimal velocity of streamers in our measurements is approximately 0.1 mm / ns = @xmath4 m / s .
the same minimal velocity has been reported for tendrils in sprites .
we also investigate the size of the initial ionization cloud at the electrode tip from which the streamers emerge , and the streamer length between branching events .
the same quantities are also measured in nitrogen with a purity of approximately 99.9% .
we characterize the essential differences with streamers in air and find a minimal diameter of @xmath5 in our nitrogen . |
hubble ( 1936 ) introduced lenticular ( s0 ) galaxies as a morphological transition class between ellipticals and early - type spirals .
the lenticular galaxies have disks with luminosity ranging from ten to hundred percent of the bulge luminosity , but without any conspicuous spiral arms . from their appearance , and also their stellar content , they seem to be more like ellipticals rather than spirals , and have often been misclassified due to this fact .
it has been suggested ( van den bergh 1994 ) that there could be different , but overlapping , sub - populations amongst the lenticulars .
there are several scenarios possible for the formation of these galaxies .
they could be of primordial origin , or could have been formed by the stripping of gas from spirals , which changes the morphology ( abadi , moore & bower 1999 ) , or through the mergers of unequal - mass spirals ( bekki 1998 ) . there could be a significant difference in the ages of disks and bulges , in which case they should have different stellar population , leading to different colors .
bothun & gregg ( 1990 ) found that bulges and disks of lenticulars were well separated in the @xmath9 vs @xmath10 diagram which they interpreted to mean that the disks are younger than bulges by 3 - 5 gyr .
but peletier & balcells ( 1996 ) found that the color differences between the disks and bulges are much smaller than those found by bothun & gregg ( 1990 ) , and inferred that the disks are only slightly younger than bulges , with the difference in age being in the range 0 - 3 gyr . in a study of 12 highly inclined lenticulars ,
michard & poulain ( 2000 ) found that disks were often redder than bulges , which indicates that there could be large concentration of dust in the disks .
a detailed multiband study of the morphology of representative samples of lenticulars , with possible separation of bulge and disk components , and comparison of their properties with those of ellipticals , and with the bulges and disks of spirals will be very important in addressing these and other possibilities .
we have observed a sample of 34 lenticular galaxies in the optical and infrared bands , and present in this paper their surface and aperture photometry , together with color profiles and mean color gradients .
we will in a subsequent publication present a decomposition of the surface brightness distribution of the galaxies into bulge and disk components using the decomposition technique developed by wadadekar , robbason & kembhavi ( 1999 ) . using the results of the decomposition
, it should be possible to see where the lenticulars are situated relative to the fundamental plane of ellipticals ( see jorgensen , franx & kjaergaard 1996 and references therein ) and the photometric plane for ellipticals and the bulges of spiral galaxies ( khosroshahi , wadadekar & kembhavi 2000a and 2000b ) . with multiband data , it might be possible to determine in this manner the bulge - to - disk luminosity ratios of these galaxies in different bands , as well as the distribution of different colors as a function of distance from the center in the bulge and disk components separately , thus facilitating the comparison with other galaxy types
. our data set will be very useful for detailed comparison with other galaxy types , as well as in constraining stellar population models for the lenticulars . the present paper deals with the details of the sample selection , observations , data reduction and analysis .
this paper is organized as follows : the selection criteria for the sample , details of observations and the technique used in data reduction are described in @xmath112 . the analysis and results from the surface photometry and aperture photometry are discussed in @xmath113 .
the distribution of color gradients are discussed in @xmath114 and a summary of the results is presented in @xmath115 .
an appendix contains comments on individual galaxies .
ccccccc ugc 00080 & ngc 0016 & @xmath123.0 & 1.8x1.0 & 0.010194 & 12.50 & @xmath1220.93 + ugc 00491 & ngc 0252 & @xmath121.0 & 1.5x1.1 & 0.016645 & 13.40 & @xmath1221.65 + ugc 00859 & ngc 0473 & 0.0 & 1.7x1.1 & 0.007118 & 13.20 & @xmath1219.82 + ugc 00926 & ngc 0499 & @xmath122.5 & 1.6x1.3 & 0.014673 & 13.00 & @xmath1221.55 + ugc 01250 & ngc 0670 & @xmath122.0 & 2.0x1.0 & 0.012352 & 13.10 & @xmath1220.76 + ugc 01823 & ngc 0890 & @xmath123.0 & 2.5x1.7 & 0.013323 & 12.50 & @xmath1222.31 + ugc 01964 & ngc 0940 & @xmath122.0 & 1.2x1.0 & 0.017319 & 13.40 & @xmath1221.64 + ugc 02039 & ngc 0969 & @xmath122.0 & 1.7x1.6 & 0.015054 & 13.50 & @xmath1221.51 + ugc 02187 & ngc 1040 & @xmath122.0 & 1.7x0.8 & 0.016054 & 14.00 & @xmath1221.06 + ugc 02322 & ngc 1106 & @xmath121.0 & 1.8x1.8 & 0.014467 & 13.70 & @xmath1221.39 + ugc 03087 & & @xmath122.0 & 0.8x0.6 & 0.033010 & 14.20 & @xmath1222.28 + ugc 03178 & ngc 1671 & @xmath122.0 & 1.1x0.9 & 0.021270 & 13.90 & @xmath1221.63 + ugc 03452 & ngc 2208 & @xmath122.0 & 1.7x1.0 & 0.018763 & 14.00 & @xmath1221.26 + ugc 03536 & & @xmath122.0 & 1.1x0.6 & 0.015641 & 14.40 & @xmath1220.86 + ugc 03567 & & @xmath122.0 & 1.0x0.8 & 0.020164 & 14.20 & @xmath1221.26 + ugc 03642 & & @xmath122.0 & 1.5x1.1 & 0.015004 & 13.50 & @xmath1221.47 + ugc 03683 & & @xmath122.0 & 2.0x1.3 & 0.019076 & 14.10 & @xmath1221.53 + ugc 03699 & ngc 2332 & @xmath122.0 & 1.5x1.0 & 0.019467 & 14.00 & @xmath1221.52 + ugc 03792 & & 0.0 & 1.8x1.3 & 0.020608 & 14.00 & @xmath1221.46 + ugc 03824 & & @xmath122.0 & 0.9x0.8 & 0.017866 & 14.40 & @xmath1220.76 + ugc 04347 & ngc 2563 & @xmath122.0 & 2.1x1.5 & 0.014944 & 13.70 & @xmath1221.06 + ugc 04767 & & @xmath122.0 & 1.3x1.1 & 0.024127 & 14.00 & @xmath1221.80 + ugc 04901 & ngc 2804 & @xmath122.0 & 2.2x2.0 & 0.028099 & 14.00 & @xmath1222.23 + ugc 05292 & ngc 3032 & @xmath122.0 & 2.0x1.8 & 0.005114 & 13.00 & @xmath1219.43 + ugc 06013 & & @xmath122.0 & 0.9x0.8 & 0.021898 & 13.90 & @xmath1221.69
+ ugc 06389 & ngc 3648 & @xmath122.0 & 1.3x0.8 & 0.006631 & 13.50 & @xmath1219.50 + ugc 06899 & ngc 3971 & @xmath122.0 & 1.4x1.2 & 0.022489 & 13.90 & @xmath1221.75 + ugc 07142 & ngc 4143 & @xmath122.0 & 2.3x1.4 & 0.003616 & 12.00 & @xmath1220.03 + ugc 07473 & ngc 4350 & @xmath122.0 & 3.0x1.4 & 0.004140 & 11.50 & @xmath1220.48 + ugc 07880 & ngc 4638 & @xmath123.0 & 2.2x1.4 & 0.003883 & 12.20 & @xmath1219.64 + ugc 07933 & ngc 4673 & @xmath125.0 & 1.0x0.9 & 0.022856 & 13.70 & @xmath1221.99 + ugc 08675 & ngc 5273 & @xmath122.0 & 2.8x2.5 & 0.003616 & 12.70 & @xmath1219.24 + ugc 09200 & ngc 5580 & @xmath122.0 & 1.8x1.8 & 0.010814 & 13.60 & @xmath1220.46 + ugc 09592 & ngc 5784 & @xmath122.0 & 1.9x1.8 & 0.017912 & 13.70 & @xmath1221.78 + ugc 11178 & ngc 6599 & @xmath122.0 & 1.3x1.2 & 0.010120 & 13.70 & @xmath1220.30 + ugc 11356 & ngc 6703 & @xmath122.5 & 2.5x2.3 & 0.008209 & 12.40 & @xmath1221.14 + ugc 11781 & & @xmath122.0 & 1.4x1.1 & 0.015501 & 13.70 & @xmath1221.74 + ugc
11972 & ngc 7248 & @xmath122.5 & 1.7x0.9 & 0.014627 & 13.60 & @xmath1221.22 + ugc 12443 & ngc 7539 & @xmath122.0 & 1.5x1.2 & 0.020174 & 13.70 & @xmath1221.88 + ugc 12655 & & @xmath122.0 & 1.4x0.9 & 0.017269 & 14.00 & @xmath1221.24
our sample consists of 40 bright and medium sized galaxies , classified as lenticulars , from the uppsala general catalogue of galaxies ( ugc ) .
the ugc is essentially complete to a limiting major - axis diameter of one @xmath13 and/or to a limiting apparent magnitude of 14.5 on the blue prints of the palomar observatory sky survey ( poss ) .
coverage is limited to the sky north of declination @xmath14 . for our sample
we selected galaxies with apparent blue magnitude brighter than @xmath15 , diameter @xmath16 and declination in the range @xmath17 .
the sample galaxies are listed in table 1 . our sample , though not complete , is representative of lenticular galaxies in the field with the above properties .
our optical @xmath18-band observations were carried out with the _ observatorio astrofsico guillermo haro _ 2.1-m telescope at cananea , mexico .
we used a tektronics ccd of @xmath19 pixel format at the @xmath20 cassegrain focus of the telescope with @xmath21 pixel binning , resulting in a scale of @xmath22 pixel@xmath23 and a field of view of @xmath24 .
typically two exposures each of ten , five and five minutes in @xmath25 , and @xmath26 bands respectively were taken for each galaxy .
exposure times were reduced if the central pixels were close to saturation .
twilight sky exposures were taken for flat - fielding purposes .
several bias frames were obtained at the start and end of each night .
all the observations were carried out in three runs in december 2001 , february 2002 and october 2002 .
a total of 34 galaxies were observed in the optical bands . for the galaxies ugc1964 , 3178 , 4901 , 6013 and 7933 from the observed
set , the quality of the images was not good enough for photometric analysis .
table 2 contains a detailed log of the observations .
we obtained images of all 40 sample galaxies in the near infrared @xmath1 band with the _ observatorio astronomico nacional _
2.1-m telescope at san pedro martir , mexico .
the camila instrument ( cruz - gonzalez 1994 ) , which hosts a nicmos 3 detector of 256@xmath27256 pixel format , was used in the imaging mode with the focal reducer configuration @xmath28 in all our observations .
this results in a spatial resolution of @xmath29 pixel@xmath23 and a total field of view of @xmath30 .
each @xmath1 observation consisted of a sequence of object and sky exposures , with the integration time of an individual exposure limited by the sky counts ( or in some cases the nucleus ) , which was kept well below the non - linear regime of the detector . a typical @xmath1 image sequence consisted of 10 exposures , six on the object and four on the sky .
the net exposure times were typically of 10 minutes .
a series of twilight and night - sky images were taken for flat - fielding purposes .
the @xmath1 observations were carried out in four runs in december 2000 , march 2001 , october 2001 and march 2002 .
the sky conditions for both optical and near infrared observations were generally photometric and the seeing _
fwhm _ was in the range 1@xmath312@xmath31 on different nights .
the average sky brightness was 21.14 , 20.72 , 20.18 and 12.27 magnitude arcsec@xmath32 in the @xmath33 and @xmath1 bands , respectively .
the sky brightness in the @xmath1 band also includes the background emitted by the warm optics .
ccccccccccc ugc 00080 & 04 - 10 - 2002 & 10 - 10 - 2001 & 1200 & 360 & 540 & 750 & 2.24 & 2.07 & 2.03 & 1.56 + ugc 00491 & 05 - 10 - 2002 & 12 - 10 - 2001 & 1200 & 600 & 600 & 750 & 1.85 & 2.18 & 2.18 & 1.52 + ugc 00859 & 14 - 12 - 2001 & 09 - 10 - 2001 & 1200 & 600 & 600 & 900 & 2.41 & 2.81 & 2.64 & 1.76 + ugc 00926 & 05 - 10 - 2002 & 10 - 10 - 2001 & 1200 & 360 & 540 & 750 & 1.80 & 1.66 & 1.78 & 1.49 + ugc 01250 & 04 - 10 - 2002 & 10 - 10 - 2001 & 1200 & 600 & 600 & 900 & 2.31 & 2.10 & 2.01 & 1.48 + ugc 01823 & 14 - 12 - 2001 & 09 - 10 - 2001 & 1200 & 600 & 600 & 750 & 2.43 & 2.78 & 2.67 & 1.61 + ugc
01964 & 14 - 12 - 2001 & 09 - 10 - 2001 & 1200 & 600 & 600 & 900 & 2.10 & 2.09 & 2.18 & 1.60 + ugc 02039 & 05 - 10 - 2002 & 12 - 10 - 2001 & 1200 & 600 & 600 & 600 & 1.91 & 2.11 & 2.01 & 1.64 + ugc 02187 & ... & 10 - 10 - 2001 & .... & ... & ... & 900 & ... & ... & ... & 1.47 + ugc 02322 & ... & 09 - 10 - 2001 & .... & ... & ... & 780 & ... & ... & ... & 1.49 + ugc 03087 & 08 - 02 - 2002 & 24 - 03 - 2002 & 1200 & 540 & 540 & 480 & 2.16 & 1.98 & 2.14 & 2.06 + ugc 03178 & 09 - 02 - 2002 & 23 - 03 - 2002 & 1200 & 600 & 600 & 320 & 2.02 & 1.87 & 1.87 & 1.75 + ugc 03452 & 14 - 12 - 2001 & 16 - 12 - 2000 & 1200 & 600 & 600 & 600 & 2.55 & 2.76 & 2.63 & 1.58 + ugc 03536 & 08 - 02 - 2002 & 26 - 03 - 2002 & 1200 & 540 & 540 & 105 & 2.14 & 2.01 & 2.05 & 1.54 + ugc 03567 & 07 - 02 - 2002 & 28 - 02 - 2002 & 1200 & 600 & 600 & 750 & 2.17 & 1.97 & 2.04 & 1.92 + ugc 03642 & 09 - 02 - 2002 & 26 - 02 - 2002 & 1200 & 720 & 720 & 750 & 1.86 & 2.04 & 1.90 & 1.48 + ugc 03683 & 08 - 02 - 2002 & 23 - 03 - 2002 & 1200 & 600 & 600 & 750 & 2.26 & 2.26 & 1.97 & 2.07 + ugc 03699 & 05 - 10 - 2002 & 12 - 10 - 2001 & 600 & 360 & 540 & 750 & 2.58 & 2.58 & 2.59 & 1.89 + ugc 03792 & 14 - 12 - 2001 & 10 - 10 - 2001 & 1200 & 600 & 600 & 750 & 2.28 & 2.69 & 2.43 & 1.85 + ugc 03824 & 07 - 02 - 2002 & 26 - 03 - 2002 & 1800 & 600 & 600 & 105 & 2.13 & 2.13 & 2.24 & 1.66 + ugc 04347 & 14 - 12 - 2001 & 08 - 03 - 2001 & 1200 & 600 & 600 & 540 & 2.51 & 2.71 & 2.55 & 1.88 + ugc 04767 & 08 - 02 - 2002 & 24 - 03 - 2002 & 1200 & 600 & 600 & 600 & 2.42 & 2.07 & 2.08 & 1.92 + ugc 04901 & 09 - 02 - 2002 & 26 - 03 - 2002 & .... & 600 & 600 & 105 & ... & 1.75 & 1.85 & 1.45 + ugc 05292 & ... & 07 - 06 - 2001 & .... & ... & ... & 900 & ... & ... & ... & 1.67 + ugc 06013 & 07 - 02 - 2002 & 08 - 02 - 2001 & .... & 600 & 600 & 840 & ... & 2.50 & 2.07 & 1.64 + ugc 06389 & 08 - 02 - 2002 & 23 - 03 - 2002 & 1200 & 540 & 540 & 700 & 1.92 & 2.05 & 2.03 & 1.89 + ugc 06899 & 08 - 02 - 2002 & 23 - 03 - 2002 & 1200 & 600 & 600 & 840 & 1.96 & 1.97 & 1.94 & 1.27 + ugc 07142 & ... & 23 - 03 - 2002 & .... & ... & ... & 450 & ... & ... & ... & 2.07 + ugc 07473 & 09 - 02 - 2002 & 22 - 03 - 2002 & 900 & 450 & 270 & 260 & 2.12 & 2.13 & 2.08 & 1.43 + ugc 07880 & 09 - 02 - 2002 & 25 - 03 - 2002 & 1200 & 270 & 600 & 500 & 2.37 & 2.64 & 2.41 & 1.79 + ugc 07933 & 07 - 02 - 2002 & 08 - 03 - 2002 & .... & 540 & 540 & 840 & ... & 2.47 & 2.26 & 1.79 + ugc 08675 & 08 - 02 - 2002 & 22 - 03 - 2002 & 1200 & 600 & 540 & 310 & 2.04 & 1.91 & 1.82 & 1.63 + ugc 09200 & ... & 07 - 03 - 2001 & .... & ... & ... & 840 & ... & ... & ... & 1.50 + ugc 09592 & ... & 24 - 03 - 2002 & .... & ... & .... & 600 & ... & ... & ... & 1.92 + ugc 11178 & 04 - 10 - 2002 & 22 - 03 - 2002 & 900 & 600 & 600 & 240 & 2.15 & 2.06 & 2.20 & 1.39 + ugc 11356 & 05 - 10 - 2002 & 23 - 03 - 2002 & 1200 & 240 & 360 & 420 & 1.91 & 1.91 & 1.91 & 1.80 + ugc 11781 & 04 - 10 - 2002 & 09 - 10 - 2001 & 900 & 600 & 600 & 540 & 1.99 & 1.94 & 2.07 & 1.83 + ugc 11972 & 05 - 10 - 2002 & 10 - 10 - 2001 & 900 & 540 & 360 & 900 & 2.07 & 1.79 & 1.78 & 1.50 + ugc 12443 & 14 - 12 - 2001 & 12 - 10 - 2001 & 1200 & 600 & 600 & 900 & 2.87 & 2.89 & 2.92 & 1.56 + ugc 12655 & 04 - 10 - 2002 & 09 - 10 - 2001 & 1200 & 600 & 600 & 120 & 2.24 & 1.98 & 2.12 & 1.59 the basic data reduction for both the optical and near infrared @xmath1 images involved subtraction of the bias and sky frames , division by flat field frames , registration of the images to a common co - ordinate system and then stacking all the images of a given galaxy in each filter .
night to night variations of the optical bias frames were negligible , and hence bias frames of an entire run were stacked together using the median algorithm to form a master bias frame , which was then subtracted from all the other frames .
preparation of the optical flat fields followed the conventional technique , wherein bias subtracted flats were stacked and the resultant frame was normalized to its mean value to form a master flat in each filter .
bias subtracted images of the program galaxies were divided by the normalized flat field in the corresponding filter .
the optical images suffered from a stray light problem that resulted in a gradient in the sky background , which roughly ran through one of the diagonals of the ccd chip .
the gradient was found to be stable throughout each run and the mean counts scaled linearly with exposure time .
after several experiments , we found that the best way to get rid of the gradient was to subtract a mean blank sky image from the data images .
for this purpose , special blank fields were observed in each filter with exposure times matching the typical exposure times of the object frames . in the december 2001 run
, three blank sky frames were used in each filter to prepare a median gradient image , while in the february 2002 run , six frames were used . the adopted procedure eliminated any gradient from the sky background . for the @xmath1 images , a bias frame taken immediately before an object exposure was subtracted as part of the data acquisition .
a master @xmath1 flat field for each night of observing was prepared as follows .
the night - sky flats were first stacked and then subtracted from stacked twilight flats .
the frames obtained in this fashion for each run were then combined and normalized to the mean value of the resultant frame to form a master @xmath1 flat .
the sky frames of each sequence of observations were combined and the resultant image was subtracted from each of the object frames to get a sky - subtracted image .
flat fielding was done by dividing the sky subtracted images of the object by the normalized master flat .
the resulting images were aligned to a common co - ordinate system using common stars in the frames and then combined using the median operation . only good images ( as defined in the camila manual
see cruz - gonzalez et al .
1994 ) were used in the @xmath1 combination .
the resulting combined @xmath1 images were aligned to corresponding images from the digitized sky survey ( dss ) . as a final step of the reduction procedure
, the mutually aligned optical images were aligned to the @xmath1 image coordinate system .
the transformed star positions in the images agreed to within @xmath34 as judged from the coordinates of common stars .
all image reductions were carried out using the image reduction and analysis facility ( iraf ) and the space telescope science data analysis system ( stsdas ) .
the iraf external package _ color _ was used to make color composite images , by combining images taken in three different bands , @xmath35 and @xmath26 or @xmath36 and @xmath1 .
gray - scale representation of these color composite images are shown in fig .
1 , where north is up and east is to the left .
dipper asterism stars in the m67 field were observed to enable accurate photometric calibration of our optical observations .
the stars in this field span a wide color range ( @xmath37 ) , which includes the range of colors of the program galaxies , and hence are suitable for obtaining the transformation coefficients to the cousins @xmath18 system defined by bessell ( 1990 ) . the transformation equations are @xmath38 where @xmath35 and @xmath26 are standard magnitudes , @xmath39 and @xmath40 are extinction corrected instrumental magnitudes , @xmath41 and @xmath42 are the zero points and @xmath43 and @xmath44 the color coefficients in bands @xmath35 and @xmath26 respectively .
typical extinction coefficients for the observatory ( 0.20 , 0.11 and 0.07 for @xmath35 and @xmath26 bands respectively ) were used . considering that the objects and the standard stars were observed as close to the meridian as possible , and in none of the cases the airmass exceeded 1.3 , the error introduced due to possible variation in the extinction coefficients is less than 0.02 magnitude .
coefficients @xmath45 and @xmath46 were obtained by using the @xmath18 standard magnitudes of chevalier & ilovaisky ( 1991 ) .
resulting values of @xmath45 and @xmath46 are given in table 3 .
the relatively large color coefficient on the @xmath26-band calibration is due to the non - standard nature of the filter used in the observations .
the stability of the @xmath45 on a given night was checked using at least two other standard fields from the landolt selected areas ( landolt 1992 ) .
the standard fields observed during our runs are sa110@xmath12232 , pg2336 + 004 , rubin149 and pg1323@xmath12086 .
overall , the zero points within a single night agreed to within 0.02 magnitudes .
night - to - night variation of the zero point was also within 0.02 magnitudes .
the detector and filter system combination that we used for the near infrared observations is identical to that used in the observations of standards by hunt et al .
( 1998 ) , and hence the color coefficients are expected to be negligibly small .
we verified this by observing fields as17 and as36 , which contain stars spanning a wide range of colors .
we observed at least 2 standard fields each night , each field containing more than one star and some fields such as as17 containing 5 stars .
the @xmath1-band zeropoints was obtained for each night as the mean of zero points from different stars and the average value is @xmath47 = @xmath48 . for reddening corrections due to galactic extinction we adopt the values given by schlegal ( 1998 ) for each filter in the optical range . for near infrared observations we use the extinction law given by rieke ( 1985 ) .
when galaxies at different redshift are seen through a fixed passband , the collected light comes from different wavelength ranges in the rest frame of the respective galaxies .
@xmath49-correction is used to transform the rest frame magnitudes of each object to the passband of the filter .
we have used an approximate form of the @xmath49-correction , applicable for small @xmath50 , from persson ( 1979 ) , frei ( 1994 ) and fukugita ( 1995 ) . for the four filters we have , @xmath51 , @xmath52 , @xmath53 and @xmath54 respectively , where @xmath50 is the redshift of the galaxy .
we have listed the combination of galactic extinction and @xmath49-correction for individual galaxies in table 4 .
ccc february 2002 & & + + @xmath55 & [email protected] & @[email protected] + @xmath57 & [email protected] & [email protected] + @xmath26 & [email protected] & @[email protected] + + + october 2002 & & + + @xmath55 & [email protected] & @[email protected] + @xmath57 & [email protected] & [email protected] + @xmath26 & [email protected] & @[email protected]
an accurate estimation of the sky background is a crucial step in surface photometric analysis , as even small uncertainties in determining the sky can lead to significant errors in the estimation of the galaxy surface brightness and color profiles , especially where the surface brightness becomes comparable to or less than the sky brightness . in those cases where the galaxy image is small enough that the frame contains portions of the sky unaffected by the galaxy , it is possible to estimate the sky using the `` boxes method '' ( see peletier 1990a ) , where the background is estimated from a series of boxes chosen avoiding the sample galaxy , foreground stars and any other contaminating objects which may be present .
we have used 20 boxes of size @xmath58 pixels near the corners of the ccd frame , and have adopted the mean of the median count in these boxes as the sky background .
the rms dispersion over the mean of 20 boxes is found to be 0.6% of the mean in the @xmath35 , and @xmath26 bands and 0.2% of the mean in the @xmath1 band .
when a galaxy is so large that the whole frame is affected by it , use of the boxes method leads to overestimation of the background . in such cases ,
the background can be determined by fitting a power - law of the form i(r ) = sky + i(0 ) r@xmath59 to the outer parts of the surface brightness profile . following jorgensen ( 1992 ) and goudfrooij ( 1994 )
, we fit the function to the region of the profile with @xmath60 typically , separately for @xmath61 and 3 , and use the mean of the respective best - fit constant values as the background .
the error in the sky background estimation is taken to be half the difference of the sky values obtained for the two fits .
we have used both the techniques for all galaxies in our sample , and find that in most cases the two methods agree within @xmath62 magnitude ( i.e. 1@xmath63 difference ) , because of the small size of the galaxies relative to the ccd frame . in the case of the galaxies ugc4347 , 7473 , 7880 , 8675 and ugc11356 , which cover most of the frame , the power - law fit provides a background which is @xmath64 magnitude ( i.e. 10@xmath63 difference ) fainter than the boxes method , and we adopt the former as the background estimation . in the @xmath1 band images , the background is already subtracted from the many individual frames for each galaxy during the pre - processing stage ( see @xmath65 ) . however , to check for any residual background which may be present in processed frames used in the analysis , we have again applied the two techniques for the background estimation .
we find that the two estimates agree very well ( @xmath66 magnitude , except in two cases where the agreement is within @xmath67 magnitude ) .
ccccccccc ugc 00080 & [email protected] & [email protected] & [email protected] & [email protected] & 0.25 & 0.06 & 0.10 & 0.20 + ugc 00491 & [email protected] & [email protected] & [email protected] & [email protected] & 0.32 & 0.07 & 0.13 & 0.27 + ugc 00859 & [email protected] & [email protected] & [email protected] & [email protected] & 0.41 & 0.09 & 0.16 & 0.30 + ugc 00926 & [email protected] & [email protected] & [email protected] & [email protected] & 0.36 & 0.08 & 0.14 & 0.29 + ugc 01250 & [email protected] & [email protected] & [email protected] & [email protected] & 0.36 & 0.08 & 0.14 & 0.29 + ugc 01823 & [email protected] & [email protected] & [email protected] & [email protected] & 0.39 & 0.09 &
0.15 & 0.31 + ugc 02039 & [email protected] & [email protected] & [email protected] & [email protected] & 0.57 & 0.13 & 0.22 & 0.44 + ugc 03087 & [email protected] & [email protected] & [email protected] & [email protected] & 0.42 & 0.10 & 0.17 & 0.35 + ugc 03452 & [email protected] & [email protected] & [email protected] & [email protected] & 0.74 & 0.17 & 0.29 & 0.56 + ugc 03536 & [email protected] & [email protected] & [email protected] & [email protected] & 0.54 & 0.12 & 0.21 & 0.42 + ugc 03567 & [email protected] & [email protected] & [email protected] & [email protected] & 0.44 & 0.10 & 0.18 & 0.36 + ugc 03642 & [email protected] & [email protected] & [email protected] & [email protected] & 0.26 & 0.06 & 0.11 & 0.23 + ugc 03683 & [email protected] & [email protected] & [email protected] & [email protected] & 0.48 & 0.11 & 0.19 & 0.39 + ugc 03699 & [email protected] & [email protected] & [email protected] & [email protected] & 0.45 & 0.10 & 0.18 & 0.37 + ugc 03792 & [email protected] & [email protected] & [email protected] & [email protected] & 0.38 & 0.09 & 0.15 & 0.32 + ugc 03824 & [email protected] & [email protected] & [email protected] & [email protected] & 0.31 & 0.07 & 0.13 & 0.27 + ugc 04347 & [email protected] & [email protected] & [email protected] & [email protected] & 0.25 & 0.06 & 0.10 & 0.22 + ugc 04767 & [email protected] & [email protected] & [email protected] & [email protected] & 0.19 & 0.04 & 0.08 & 0.21 + ugc 06389 & [email protected] & [email protected] & [email protected] & [email protected] & 0.11 & 0.02 & 0.04 & 0.10 + ugc 06899 & [email protected] & [email protected] & [email protected] & [email protected] & 0.17 & 0.04 & 0.08 & 0.19 + ugc 07473 & [email protected] & [email protected] & [email protected] & [email protected] & 0.13 & 0.03 & 0.05 & 0.10 + ugc 07880 & [email protected] & [email protected] & [email protected] & [email protected] & 0.12 & 0.03 & 0.05 & 0.09 + ugc 08675 & [email protected] & [email protected] & [email protected] & [email protected] & 0.05 & 0.01 & 0.02 & 0.05 + ugc 11178 & [email protected] & [email protected] & [email protected] & [email protected] & 0.70 & 0.16 & 0.27 & 0.51 + ugc 11356 & [email protected] & [email protected] & [email protected] & [email protected] & 0.41 & 0.09 & 0.16 & 0.31 + ugc 11781 & [email protected] & [email protected] & [email protected] & [email protected] & 1.12 & 0.26 & 0.43 & 0.82 + ugc 11972 & [email protected] & [email protected] & [email protected] & [email protected] & 0.76 & 0.17 & 0.29 & 0.56 + ugc 12443 & [email protected] & [email protected] & [email protected] & [email protected] & 0.52 & 0.12 & 0.21 & 0.42 + ugc 12655 & [email protected] & [email protected] & [email protected] & [email protected] & 0.32 & 0.07 & 0.13 & 0.27 we have fitted ellipses to the isophotes in our @xmath68 and @xmath1 band images , using the task ellipse in the stsdas package available in iraf . the fitting algorithm used is described in detail by jedrzejewski ( 1987 ) , and uses the intensity distribution along a trial ellipse , which can be expressed as a fourier series @xmath69 where @xmath70 is the ellipse eccentric anomaly , @xmath71 is the mean intensity along the ellipse and @xmath72 , @xmath73 are harmonic amplitudes . the fitting was started a few arcseconds from the center of the galaxy image to minimize the effect of seeing , and stopped at the isophotes where the mean count becomes comparable to three times the error in the sky background .
stars , bad pixels etc . were identified in the first round of the fitting procedure and masked in the next run .
all the parameters including the center of the ellipse were allowed to vary during the fitting .
variation in ellipse center was found to be small ( @xmath74 ) , which provides a check on the accuracy of the sky subtraction .
we repeated the fitting process with different modes of sampling and different starting major - axis lengths to check the stability of the extracted parameters .
the fitting procedure provides the mean intensity ( surface brightness ) , the ellipticity and the position angle of the major - axis of the best fit ellipse , as a function of the semi - major axis length . the @xmath55 band surface brightness profile , and the @xmath75 and @xmath5 color profiles are shown for all the galaxies in fig .
the profiles of the ellipticity , position angle and the fourier coefficients @xmath76 respectively in the @xmath77 and @xmath1 bands for one galaxy ( ugc7880 ) are shown in fig .
tables and plots of surface brightness and color profiles as well as the profiles of the ellipticity , position angle and the fourier coefficients @xmath78 respectively in the @xmath77 and @xmath1 bands for all the galaxies in the sample are available at _ http://iucaa.ernet.in/@xmath79sudhan / s0.html_. some description of the profiles for individual galaxies is provided in the appendix .
color measurements of galaxies involving frames with different seeing full width at half maximum ( _ fwhm _ ) can lead to errors at small radii , and in particular within the seeing disk which has to be approached for measuring nuclear colors .
these errors have been discussed by franx ( 1989 ) , peletier ( 1990a ) and goudfrooij ( 1994 ) , who have derived a cutoff radius for color profiles and discarded colors measured inside this limit , which is usually @xmath80 .
idiart ( 2002 ) used a method of equalization of the _ fwhms _ to reduce error due to different seeing _ fwhm _ but this method is feasible only if the frames involved in color measurement are observed in quick succession . for our observations
the seeing _
fwhm _ was nearly the same for the optical @xmath35 and @xmath26 bands , but significantly better in @xmath1 band as indicated by table 2 .
we have therefore degraded the @xmath1 images to the mean _ fwhm _ of the optical band images for each galaxy .
the @xmath5 color profiles shown in fig .
2 were obtained using the degraded @xmath1 band images .
cccccc ugc 00080 & 3.40 & 28.05 & @[email protected] & @[email protected] & @[email protected] + ugc 00491 & 3.40 & 51.00 & @[email protected] & @[email protected] & @[email protected] + ugc 00859 & .... & ..... & ..... & ..... & ..... + ugc 00926 & 3.40 & 34.00 & @[email protected] & @[email protected] & @[email protected] + ugc 01250 & .... & .... & ..... & ..... & @[email protected] + ugc 01823 & 3.40 & 34.00 & @[email protected] & @[email protected] & @[email protected] + ugc 02039 & 3.40 & 09.35 & @[email protected] & @[email protected] & @[email protected] + ugc 03087 & .... & .... & ..... & ..... & ..... + ugc 03452 & 3.40 & 09.35 & @[email protected] & @[email protected] & @[email protected] + ugc 03536 & 3.40 & 09.35 & @[email protected] & @[email protected] & @[email protected] + ugc 03567 & 3.40 & 09.35 & @[email protected] & @[email protected] & @[email protected] + ugc 03642 & 3.40 & 13.60 & @[email protected] & @[email protected] & @[email protected] + ugc 03683 & 3.40 & 13.60 & @[email protected] & @[email protected] & @[email protected] + ugc 03699 & 3.40 & 17.00 & @[email protected] & @[email protected] & @[email protected] + ugc 03792 & 3.40 & 17.00 & @[email protected] & @[email protected] & @[email protected] + ugc 03824 & 3.40 & 21.25 & @[email protected] & @[email protected] & @[email protected] + ugc 04347 & 3.40 & 22.95 & @[email protected] & @[email protected] & @[email protected] + ugc 04767 & 3.40 & 18.70 & @[email protected] & @[email protected] & [email protected] + ugc 06389 & 3.40 & 22.95 & @[email protected] & @[email protected] & @[email protected] + ugc 06899 & 3.40 & 17.85 & @[email protected] & @[email protected] & @[email protected] + ugc 07473 & 3.40 & 17.00 & @[email protected] & @[email protected] & @[email protected] + ugc 07880 & 3.40 & 16.15 & @[email protected] & @[email protected] & @[email protected] + ugc 08675 & 5.95 & 53.55 & @[email protected] & @[email protected] & @[email protected] + ugc 11178 & 3.40 & 17.85 & @[email protected] & @[email protected] & @[email protected] + ugc 11356 & 3.40 & 21.25 & @[email protected] & @[email protected] & @[email protected] + ugc 11781 & 3.40 & 08.50 & @[email protected] & @[email protected] & @[email protected] + ugc 11972 & 3.40 & 19.55 & @[email protected] & @[email protected] & @[email protected] + ugc 12443 & 3.40 & 14.45 & @[email protected] & @[email protected] & @[email protected] + ugc 12655 & 3.40 & 14.45 & @[email protected] & @[email protected] & @[email protected] lccc our sample & @xmath6 ( @xmath81 ) & @xmath7 ( @xmath81 ) & @xmath8 ( @xmath81 ) + + ellipticals & @xmath82 ( @xmath83 ) & @xmath84 ( @xmath85 ) & @xmath86 ( @xmath83 ) + ( peletier 1990a & b ) & & & + + bulges of disk galaxies & ... &
@xmath87 ( @xmath85 ) & ... + ( peletier & balcells 1997 ) & & & in this work we obtain the total magnitude of each galaxy in all the bands by integrating the light within the isophote corresponding to @xmath88 = 25 magarcsec@xmath32 in each band . to obtain the @xmath75 and @xmath5 colors , however , we integrate out only to the radius at which the error in @xmath88 is 0.1 magarcsec@xmath32 ( the radius where the dotted horizontal line intersects the @xmath89 profile in fig . 2 ) .
the colors obtained in this manner are consistent with those obtained using total magnitude as defined above , within the estimated errors .
however , the errors on integrated colors obtained in this manner are substantially lower ( typically 0.01 vs 0.04 in @xmath2 , 0.01 vs 0.04 in @xmath3 , and 0.05 vs 0.12 in @xmath5 colors ) .
we also obtain total @xmath1 magnitudes measured to @xmath90 = 20 magarcsec@xmath32 in order to make comparison with 2mass photometry .
the measured total @xmath55 magnitudes and @xmath75 and @xmath5 colors along with the estimated errors for the 29 galaxies in the sample for which we have the necessary data are shown in table 4 . in fig .
4 ( left panel ) we have compared our photometry with that in the literature .
we have plotted ( from bottom to top ) the difference between ( i - ii ) our @xmath55 magnitude and @xmath2 color and those of rc3 , ( iii ) our @xmath55 magnitude and that of ugc and ( iv ) our @xmath1 magnitude measured inside the isophote @xmath91 = 20 magarcsec@xmath32 and that of 2mass , all against our @xmath55 magnitude .
agreement between our @xmath55 magnitude with that of rc3 is within the expected errors for all except two galaxies .
our total @xmath2 colors are systematically around 0.04 magnitude bluer than those of rc3 . on the other hand ,
if we compare the @xmath2 color obtained by integrating only upto @xmath89 = 21 magarcsec@xmath32 ( _ bulge color _ ) , the agreement between the colors is excellent .
the agreement between our @xmath1 magnitude is also reasonably good .
however it is important to note that 2mass magnitudes correspond to @xmath92 = 20 magarcsec@xmath32 and considerable flux could be outside this radius our total @xmath1 magnitudes ( measured inside the radius corresponding to @xmath93 = 25 magarcsec@xmath32 ) are systematically brighter by @xmath790.4 magnitudes as compared to those of 2mass .
hence @xmath1 magnitudes in the 2mass catalog underestimate the total flux . to check the internal consistency of our photometry we have plotted color - color diagrams as shown in fig .
4 ( right panel ) . in general
, there is a good correlation between these independent color measurements and except for two galaxies ugc 3087 and ugc 11178 , the estimated colors follow the sequence expected for early - type galaxies . among these ,
ugc 3087 is clearly outside the 3@xmath63 error estimation in all the three colors .
this galaxy is known to harbor an active galactic nucleus , which clearly affects the color profiles in fig .
hence its colors are expected to depart significantly from the general correlation .
ugc 11178 , though is within the 2@xmath63 error bar , is curious because it stands out in the plots , which are completely independent measurements .
further photometry would be required to understand the nature of this object .
we exclude these two galaxies from further analysis .
we have used optical and @xmath1 band data to study the spatial distribution of dust .
we have obtained extinction map in magnitude @xmath94 = @xmath95 , where @xmath96 is the observed image in same band , and @xmath97 is the smooth version of this image , obtained using the best fit ellipse parameters .
extinction map particularly helping in identifying non - axisymmetric structure .
color maps too are of help in identifying features with wavelength dependent intensity .
we have noticed from the color and extinction maps that the lenticular galaxies ugc3178 , ugc3536 , ugc3792 , ugc4347 , ugc7473 and ugc7880 show clear evidence of the presence of dust in the form of lanes or patches . in all these cases there
are differences in the ellipticity and position angle profiles , as well as in the higher order fourier coefficients profiles , in the different bands .
this indicates that such differences could be used as indicators of the possible presence of dust , even when the dust is not otherwise discernible ( see peletier 1990a , goudfrooij 1994 ) .
the lenticular galaxy ugc3792 has a very prominent dust lane along the major axis ( see the appendix for details ) .
we will present elsewhere a detailed study of the dust in our sample of galaxies .
[ fig : xxx ] ( continued ) fig .
[ fig : xxx ] ( continued ) fig .
[ fig : xxx ] ( continued ) fig .
[ fig : xxx ] ( continued )
it is the usual practice to parameterize the color gradient over the galaxy as the slope @xmath98 , which is evaluated by making a straight line fit between an inner and an outer radius to the color profiles .
the inner cutoff limit ( r1 ) is taken to be @xmath791.5 times the seeing fwhm ( peletier 1990a ) , so that seeing effects are lessened . the outer limit ( r2 )
is taken to be the point along the major axis at which the error on the mean surface brightness of the fitted ellipse reaches @xmath99 .
it is seen from fig .
2 that plots of color against the logarithm of the semi - major axis for most of our sample galaxies , have a reasonably linear form between the inner and outer cutoff .
there are some cases ( ugc859 , 1250 and 3087 ) for which the color profiles are non - linear between r1 and r2 because of the presence of a strong ring ( ugc859 , 1250 ) or an agn ( ugc3087 ) , which makes reliable estimation of the color gradient is difficult .
the @xmath75 and @xmath5 color gradients for our galaxies are given in table 5 .
the colors of the galaxies in our sample become bluer outwards , in keeping with the trend observed in ellipticals . the only exception to this
is the positive gradient in the @xmath5 color of the galaxy ugc4767 . after excluding the galaxy ugc3792 ( which has a prominent dust lane )
, the mean logarithmic gradients in the @xmath75 and @xmath5 colors are @xmath6 , @xmath7 , @xmath8 magnitude per dex in radius respectively .
a comparison of our mean color gradients with corresponding values from the literature for ellipticals and bulges of early - type spirals is shown in table 6 .
the color gradients for lenticulars are more negative than the gradients for ellipticals , while the @xmath3 gradient for lenticulars is less negative than the corresponding gradient for bulges of early - type spirals .
our steeper color gradients could imply that the metallicity gradients are stronger in lenticulars and/or that there is more recent star formation in the outer parts of lenticulars as compared to those of ellipticals .
bothun & gregg ( 1990 ) reported significant differences in the colors of bulges and disks in lenticulars which they interpret as being due to age difference , and not metallicity difference , between the bulge and disk components .
the large rms error in the mean color gradient of our sample could be due to the different colors of the bulge and disk components and varying proportion of these in different lenticulars .
correlation between different color gradients have been noted by various authors ( peletier 1990a & b and idiart 2002 ) and have been used to derive information about the physical processes in galaxies .
such correlations are displayed in fig .
5 for our sample .
a clear correlation is seen between @xmath2 and @xmath3 gradients , with correlation coefficient 0.90 at a significance level better than 99.99 percent .
correlation between @xmath2 and @xmath3 gradient has been noticed for a sample of ellipticals by peletier ( 1990 a & b ) . from their data
we obtain a correlation coefficient of 0.97 at a significance level better than 99.99 percent .
the plot of @xmath5 against @xmath2 gradients shows large scatter , and we do not see any correlation of the type reported by peletier ( 1990b ) , but it should be noted that they used only 13 galaxies to derive the correlation .
it is necessary to investigate how the separate bulge and disk color gradients affect the total color gradients and hence the correlation between them .
we will address in the future these issues using a bulge - disk decomposition technique which will allow us to study the color distribution in these components separately .
in this paper we have presented detailed multicolor surface photometry performed with a ccd in the @xmath100 bands and a nicmos detector in the @xmath1 band , for a sample of 34 lenticular galaxies from the ugc catalogue .
the galaxies were chosen in an unbiased fashion from a subset of ugc lenticulars as explained in section 2 .
we have obtained total integrated magnitudes and colors for all the galaxies using elliptical annuli from surface photometry , and find that these are in good agreement with values from the rc3 catalogue . using isophotal analysis we have obtained radial profiles of the surface brightness , ellipticity , position angle , and higher order fourier coefficients in all the bands .
the profiles in the different bands are consistent with each other , and any differences can be attributed to the presence of dust and other features which produce wavelength dependent effects . we have used the surface brightness profiles to obtain color profiles , and logarithmic color gradients , and find that the gradients are negative , indicating that the colors of lenticulars become redder towards the center , as is the case with elliptical galaxies .
we have shown that there is good correlation between @xmath2 and @xmath3 color gradients for lenticulars .
numerical profiles of all parameters that we have obtained from the isophotal analysis along with the color images are available at _ http://www.iucaa.ernet.in/@xmath79sudhan / s0.html_. our intention in obtaining the multiband data on lenticulars has been to study in detail these galaxies as a class , and to compare their properties with those of ellipticals and early type spirals .
an important aspect of this study will be the decomposition of the lenticulars into bulge and disk components and the comparison of these separately with the bulges and disks in other types of galaxies where these components occur with varying degrees of prominence .
we will use the results of the decomposition in a multiband study of the fundamental and photometric planes for lenticulars .
we will also make a detailed study of the distribution of dust in lenticulars , particularly in those galaxies from our sample where there are prominent dust features and where we have multiband data at optical as well as near - infrared wavelengths .
the staff at oagh and spm are greatfully acknowledged for their help during the observations .
sb and skp thank iucaa for hospitality and the use of facilities without which this work could not have been done and for providing funds for observations .
akk and sb thank inaoe for hospitality provided during there visits .
we also thank the anonymous referee for several useful comments , which helped to improve the original manuscript .
this research has made use of the nasa / ipac extragalactic database ( ned ) , which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration . in this appendix
, we discuss structural properties of each galaxy , whenever isophotes in one or more bands depart from a smooth ellipse .
the discussions are focused on any disagreement in the morphological classification in ugc and rc3 catalogs , presence of dust lane , change of ellipse - fit parameters ( b4 , pa or ellipticity ) , existence of any ring , and any evidence of nuclear activity .
published information on any of these issues , if available , is noted . in total 12 galaxies ( ugc926 , 1823 , 3452 , 3567 , 3683 , 3699 , 3824 , 6013 , 6389 , 6899 , 11972 and 12443 )
are smooth lenticular galaxies without any identifiable feature in the direct images , the surface brightness profiles , color or extinction maps .
comments for the remaining galaxies follow . * ugc80 :* ugc80 is a barred lenticular galaxy , having a very faint interacting companion .
the faint bar is almost perpendicular to the galaxy major axis , and is clearly evident in the isophotal profile .
the @xmath101 profile is indicative of a disky shape over the entire major - axis range explored .
the ellipticity and position angle profiles show abrupt change at @xmath102 , where there is also a kink in the brightness profile .
the @xmath3 color map of the galaxy shows a red feature near the center .
* ugc491 :* ugc491 is the brightest galaxy in a small group with ngc0258 and ngc0260 .
this galaxy is classified as s0 in the ugc catalogue and sa(r ) in the rc3 catalogue . in the optical images it indeed shows a ring ,
which is also evident in the @xmath3 and @xmath103 color maps .
ellipticity and position angle profiles show a change at @xmath104 , which may be due to the presence of the ring .
* ugc859 :* ugc859 shows an internal ring , but it is classified as s0 in the ugc catalogue .
the @xmath3 and @xmath103 color maps of the galaxy show the ring as an inhomogeneous structure .
a low surface brightness disk with some luminous spiral structure exists outside the ring .
ugc 859 has been detected by iras and has flux densities of @xmath105 at @xmath106 and @xmath107 at @xmath108 ( knapp 1989 ) .
the @xmath101 coefficient is disky towards the outer region . *
ugc1250 :* while this galaxy is classified as s0 , the @xmath3 color map of the galaxy show a very clear , but distorted ring , just around the bulge .
@xmath109 $ ] observations by pogge & eskridge ( 1993 ) show copious hii regions across the disk .
the @xmath101 profile is disky at all major - axis lengths , with the coefficient being larger in the inner region of the galaxy .
* ugc2039 :* ugc2039 is paired with ugc 2049 .
there is a rise in the ellipticity from 0.1 to 0.5 towards the outer region .
the @xmath101 coefficient is disky in the outer region , which suggests the presence of a significant disk .
the galaxy has been detected by iras at 60 and @xmath108 .
* ugc3087 :* previous studies of ugc3087 have mainly concentrated on its nuclear activity .
it is a strong radio source ( 3c120 ) and the optical spectrum suggests a seyfert 1 nucleus ( tadhunter 1993 ) .
it has a faint optical jet in the same apparent direction as the radio jet .
the optical jet is clearly visible in our images and distorts isophotal profiles in the inner region .
* ugc3178 :* a dust feature is clearly visible in the @xmath55 band extinction map of this galaxy , and is also detectable from the isophotal analysis of @xmath55 and @xmath1 band images .
the @xmath101 coefficient , position angle and ellipticity are different in @xmath55 and @xmath1 , which may be due to dust absorption in @xmath55 .
our @xmath57 and @xmath26 band images suffer from poor s / n .
* ugc3536 :* this galaxy has a significant dusty disk and the @xmath101 coefficient is disky .
the disk is also evident in the @xmath2 and @xmath3 color maps .
there is a constant rise in ellipticity for semi - major axis length @xmath110 , beyond which there is a decrease , which could be because of the disk .
* ugc3792 :* this galaxy is classified as s0 in the ugc but as sa0/a in the rc3 catalogue , and has no previous reported photometric study .
the galaxy has a prominent dust lane along the major axis which greatly affects the various profiles .
this dust lane is most prominent in the @xmath55 band .
the @xmath101 coefficient is positive in the outer region .
there is a large color gradient in @xmath75 and @xmath5 , presumably because of the dust .
a detailed study of this galaxy in the @xmath111 bands will be presented in a forthcoming paper . * ugc4347 :* our @xmath2 and @xmath103 color and extinction maps of this galaxy
reveal a large dust patch near the center .
the @xmath101 profile is disky throughout the observed region .
forbes & thomson ( 1992 ) have detected possible shells in this galaxy .
the extinction images in all bands shows faint structures , which may be responsible for the non - zero values the of @xmath101 coefficient .
* ugc4767 :* the position angle and ellipticity profile for this galaxy are different in the @xmath112 and @xmath1 bands for semi - major axis lengths @xmath113 , which could be because of a dust patch near the center of the galaxy .
the various color maps also indicate the presence of the dust patch . * ugc4901
:* our @xmath55 band image of this galaxy has poor s / n .
the color maps in @xmath114 and @xmath5 show no features in the galaxy .
the @xmath101 coefficient is positive towards the outer region , suggesting the presence of a faint disk .
* ugc7473 :* this well studied edge - on lenticular galaxy forms a pair with ngc 4340 .
the surface brightness profile of this galaxy clearly indicates the presence of bulge and disk components .
the mg@xmath115 line strength profiles along the major and minor axis of the galaxy differ dramatically and convincingly indicate the presence bulge and disk components ( fisher , franx & illingworth 1996 ) .
michard & marchal ( 1993 ) described ugc7473 as having a disk fully embedded in a spheroidal halo .
a disk of rapidly rotating gas is present within the inner @xmath116 ; this is decoupled from the stellar component ( fisher 1997 ) .
a concentration of dust in the disk has been proposed by michard & poulain ( 2000 ) .
our @xmath3 and @xmath103 color and extinction maps also reveal a clear inclined disk .
the ellipticity rises from 0.2 to 0.61 over the observed region and the @xmath101 coefficient becomes significantly positive beyond @xmath117 , which reflects the presence of a strong disk .
* ugc7880 :* this galaxy forms a pair with ngc 4635 , with a separation of at @xmath118 . the ellipticity changes from 0.1 to 0.5 upto @xmath119 , and decreases beyond that .
the @xmath101 coefficient is positive upto the point at which the ellipticity begins to decrease
. this may be due to the presence of an inner disk , the presence of which is also indicated by the @xmath3 and @xmath103 color and extinction maps .
* ugc7933 :* this galaxy is classified as s0 in the ugc , but as e1 - 2 in the rc3 catalogue .
it forms a non - interacting pair with ngc4670 at @xmath120 separation .
the @xmath55 band image has poor s / n .
the @xmath101 coefficient is positive towards the outer region .
the color and extinction maps do not indicate any structure .
* ugc8675 :* this is a little studied s0 galaxy hosting a seyfert 1.5 nucleus , and has been classified as a sa0 galaxy in the rc3 catalogue .
the isophotes of the galaxy are nearly circular and at @xmath121 the ellipticity becomes 0.15 which is consistent with the values obtained by de robertis , hayhoe & yee ( 1998 ) and pierre ( 2000 ) .
a dust absorption pattern is seen near the nucleus in the @xmath3 and @xmath103 color and extinction maps .
pierre ( 2000 ) have found from hst observations a u - shaped dust lane circling around the nucleus .
* ugc11178 :* this galaxy has no previous reported photometric study .
the extinction map show a ring like structure in all the bands .
the ellipticity abruptly changes between 5 and @xmath122 , which could be because of the ring like structure .
the @xmath101 coefficient is significantly positive upto @xmath122 .
the color maps are featureless .
* ugc11781 :* this galaxy is classified as s0 in the ugc but as sab in the rc3 catalogue .
the presence of a bar is clear in the direct as well as extinction images .
there is a dip in ellipticity profile at @xmath119 , where there is a change in the position angle profile as well .
the @xmath101 coefficient is positive in this region . | we present in this paper multicolor surface and aperture photometry in the @xmath0 and @xmath1 bands for a sample of 34 lenticular galaxies from the ugc catalogue . from surface photometric analysis
, we obtain radial profiles of surface brightness , colors , ellipticity , position angle and the fourier coefficients which describe the departure of isophotal shapes from purely elliptical form and find the presence of dust lanes , patches and ring like structure in several galaxies in the sample .
we obtain total integrated magnitudes and colors and find that these are in good agreement with the values from the rc3 catalogue .
isophotal colors are correlated with each other , following the sequence expected for early - type galaxies .
the color gradients in lenticulars are more negative than the corresponding gradients in ellipticals .
there is a good correlation between @xmath2 and @xmath3 color gradients , and the mean gradient in the @xmath4 and @xmath5 colors are @xmath6 , @xmath7 , @xmath8 magnitude per dex in radius respectively . |
a recent influx of academic monographs @xcite and popular books @xcite manifests a keen cultural and scientific interest in complex networks , which appeal to both applied and theoretical problems in national defense , sociology , epidemiology , computer science , statistics , and mathematics .
the erds rnyi random graph @xcite remains the most widely studied network model .
its simple dynamics endow it with remarkable mathematical properties , but this simplicity overpowers any ability to replicate realistic structure .
many other network models have been inspired by empirical observations .
chief among these is the _ scale - free _ phenomenon , which has garnered attention since the initial observation of power law behavior for internet statistics @xcite .
celebrated is barabsi and albert s preferential attachment model @xcite , whose dynamics are tied to the _ rich get richer _ or _ matthew effect_. citing overlooked attributes of network sampling schemes , other authors @xcite have questioned the power law s apparent ubiquity .
otherwise , watts and strogatz @xcite proposed a model that replicates milgram s _ small - world _
phenomenon @xcite , the vernacular notion of _ six degrees of separation _ in social networks .
networks arising in many practical settings are dynamic , they change with time .
consider a population @xmath0 of individuals . for each @xmath1
, let @xmath2 indicate a social relationship between @xmath3 and @xmath4 and let @xmath5 comprise the indicators for the whole population at time @xmath6 .
for example , @xmath2 can indicate whether @xmath3 and @xmath4 are co - workers , friends , or family , have communicated by phone , email , or telegraph within the last week , month , or year , or subscribe to the same religious , political , or philosophical ideology . within the narrow scope of social networks ,
the potential meanings of @xmath2 seem endless ; expanding to other disciplines , the possible interpretations grow . in sociology ,
@xmath7 records changes of social relationships in a population ; in other fields , the network dynamics reflect different phenomena and , therefore , can exhibit vastly different behaviors . in each case , @xmath7 is a time - varying network .
time - varying network models have been proposed previously in the applied statistics literature .
the _ temporal exponential random graph model _ ( tergm ) in @xcite incorporates temporal dependence into the _ exponential random graph model _ ( ergm ) .
the authors highlight select properties of the tergm , but consistency under subsampling is not among them . from the connection between sampling consistency and lack of interference
, it is no surprise that the exponential random graph model is sampling consistent only under a choking restriction on its sufficient statistics @xcite .
mccullagh @xcite argues unequivocally the importance of consistency for statistical models .
presently , no network model both meets these logical requirements and reflects empirical observations . in this paper , rather than
focus on a particular application , we discuss network modeling from first principles .
we model time - varying networks by stochastic processes with a few natural invariance properties , specifically , exchangeable , consistent markov processes .
the paper is organized as follows . in section [ section : modeling preliminaries ] , we discuss first principles for modeling time - varying networks ; in section [ section : informal description ] , we describe the rewiring process informally ; in section [ section : rewiring maps ] , we introduce the workhorse of the paper , the rewiring maps ; in sections [ section : discrete ] and [ section : exchangeable rewiring maps ] , we discuss a family of time - varying network models in discrete - time ; in section [ section : continuous ] , we extend to continuous - time ; in section [ section : poissonian structure ] , we show a poisson point process construction for the rewiring process , and we use this technique to establish the feller property ; and in section [ section : concluding remarks ] , we make some concluding remarks .
we prove some technical lemmas and theorems in section [ section : proof ] .
for now , we operate with the usual definition of a graph / network as a pair @xmath8 of vertices and edges . we delay formalities until they are needed .
let @xmath9 be a random collection of graphs indexed by @xmath10 , denoting _ time_. we may think of @xmath11 as a collection of social networks ( for the same population ) that changes as a result of social forces , for example , geographical relocation , broken relationships , new relationships , etc . , but our discussion generalizes to other applications . in practice , we can observe only a finite sample of individuals . since the population size is often unknown , we assume an infinite population so that our model only depends on known quantities .
thus , each @xmath12 is a graph with infinitely many vertices , of which we observe a finite sub - network @xmath13}_t$ ] with @xmath14 vertices .
since the vertex labels play no role , we always assume sampled graphs have vertex set @xmath15:=\{1,\ldots , n\}$ ] , where @xmath16 is the sample size , and the population graph is infinite with vertex set @xmath17 , the natural numbers .
the models we consider are _ markovian _ , _ exchangeable _ , and _
consistent_. the process @xmath11 has the _ markov property _
if , for every @xmath18 , its pre-@xmath6 and post-@xmath6 @xmath19-fields are conditionally independent given the present state @xmath12 .
put another way , the current state @xmath12 incorporates all past and present information about the process , and so the future evolution depends on @xmath20 only through @xmath12 .
it is easy to conceive of counterarguments to this assumption : in a social network , suppose there is no edge between individuals @xmath21 and @xmath22 or between @xmath23 and @xmath24 at time @xmath18 .
then , informally , and @xmath25 for the sake of illustration . ]
we expect the future ( marginal ) evolution of edges @xmath26 and @xmath25 to be identically distributed .
but if , in the past , @xmath21 and @xmath22 have been frequently connected and @xmath23 and @xmath24 have not , we might infer that the latent relationships among these individuals are different and , thus , their corresponding edges should evolve differently .
for instance , given their past behavior , we might expect that @xmath21 and @xmath22 are more likely than @xmath23 and @xmath24 to reconnect in the future . despite such counterarguments ,
the markov property is widely used and works well in practice .
generalizations to the markov property may be appropriate for specific applications , but they run the risk of overfitting . structure and changes to structure drive our study of networks .
vertex labels carry no substantive meaning other than to keep track of this structure over time ; thus , a suitable model is _ exchangeable _ , that is , its distributions are invariant under relabeling of the vertices . for a model on finite networks ( i.e. , finitely many vertices )
, exchangeability can be induced trivially by averaging uniformly over all permutations of the vertices .
but we assume an infinite population , for which the appropriate invariance is _ infinite exchangeability _ , the combination of exchangeability and consistency under subsampling ( section [ section : consistency ] ) .
unlike the finite setting , infinite exchangeability can not be imposed arbitrarily by averaging ; it must be an inherent feature of the model . for any graph with vertex set @xmath27
, there is a natural and obvious restriction to an induced subgraph with vertex set @xmath28 by removing all vertices and edges that are not fully contained in @xmath29 .
the assumption of _ markovian consistency _ , or simply _ consistency _
, for a graph - valued markov process implies that , for every @xmath30 , the restriction @xmath31}$ ] of @xmath11 to the space of graphs with vertex set @xmath15 $ ] is , itself , a markov process .
note that this property does not follow immediately from the markov assumption for @xmath11 because the restriction operation is a many - to - one function and , in general , a function of a markov process need not be markov .
also note that the behavior of the restriction @xmath31}$ ] can depend on @xmath11 through as much as its exchangeable @xmath32-field , which depends only on the `` tail '' of the process .
markovian consistency may be unjustified in some network modeling applications .
this contrasts with other combinatorial stochastic process models , for example , coalescent processes @xcite , for which consistency is induced by an inherent _ lack of interference _ in the underlying scientific phenomena .
nevertheless , if we assume the network is a sample from a larger network , then consistency permits out - of - sample statistical inference @xcite . without markovian consistency in a time - varying markov model ,
the sampled process can depend on the whole ( unobserved ) process , leaving little hope for meaningful inference .
we can envision at least two kinds of network dynamics that correspond , intuitively , to _ local _ and _ global _ structural changes .
local changes involve _ only one _ edge , global changes involve a _ positive fraction _ of edges .
we say the _ status _ of edge @xmath33 is _ on _ if there is an edge between @xmath21 and @xmath23 ; otherwise , we say the status is _
off_. a local change occurs whenever the status of exactly one edge changes , called a _ single - edge update_. an easy way to generate single - edge updates is by superposition of independent rate-1 poisson processes . for each pair
@xmath34 , we let @xmath35 be the arrival times of a rate-1 poisson point process . at each arrival time , the status of the edge between @xmath21 and @xmath23 changes ( either from ` off ' to ` on ' or the reverse ) . doing this independently for each pair results in an infinite number of changes to the network in any arbitrary time interval , but only finitely many changes within each finite subnetwork .
we call a process with this description a _ local - edge process _ ; see section [ section : local - edge ] .
a global change occurs whenever the status of a positive proportion of edges changes simultaneously . in practice ,
such an event might indicate a major external disturbance within the population , for example , spread or fear of a pandemic .
modeling such processes in continuous - time requires more preparation than the local - edge process . for an example , consider generating a discrete - time markov chain @xmath36 on the finite space of graphs with vertex set @xmath15 $ ] . at any time @xmath37 ,
given @xmath38 , we can generate a transition to a new state @xmath39 as follows . independently for each pair @xmath34 , we flip a coin to determine whether to put an edge between @xmath21 and @xmath23 in @xmath39 : if @xmath33 is _ on _ in @xmath40 , we flip a @xmath41-coin ; otherwise , we flip a @xmath42-coin .
this description results in a simple , exchangeable markov chain on finite graphs , which we call the _ erds rnyi rewiring chain _ ( section [ section : er ] ) .
more general transitions are possible , for example , edges need not evolve independently .
we use the next markov chain as a running example of a discrete - time rewiring chain .
we fix @xmath30 and regard an undirected graph @xmath40 with vertex set @xmath15 $ ] as a @xmath43-valued symmetric matrix @xmath44 such that @xmath45 for all @xmath46 ; that is , we represent a graph by its adjacency matrix with @xmath47 . for any pair of graphs @xmath48
, we can compute the statistic @xmath49 , where for @xmath50 , @xmath51 for example , @xmath52 is the number of pairs @xmath53 for which the status of @xmath33 changes from @xmath54 to @xmath55 from @xmath40 to @xmath39 .
we use @xmath56 as a sufficient statistic to define the transition probability @xmath57 where @xmath58 and @xmath59 .
the sufficient statistic @xmath56 is invariant under joint relabeling of the vertices of @xmath48 and so the transition law is exchangeable .
furthermore , @xmath60 is reversible with respect to @xmath61 where @xmath62 , @xmath63 .
the distribution @xmath64 arises as a mixture of erds rnyi random graphs with respect to the @xmath65 distribution .
furthermore , @xmath66 is a consistent collection of transition probabilities and , therefore , determines a unique transition probability ( and hence markov chain ) on the space of infinite graphs with vertex set @xmath67 .
though consistency is not immediately obvious for the above family , the savvy reader might anticipate it : the formula for @xmath68 involves rising factorials ( i.e. , gamma functions ) , which also appear in other consistent combinatorial stochastic processes , for example , the chinese restaurant process @xcite and the beta - splitting model for fragmentation trees @xcite .
we need not prove consistency explicitly for this model ; it follows from our more general construction of rewiring processes , all of which are consistent ( theorem [ thm : consistent rewiring chain ] ) .
we discuss the above family further in section [ section : er ] . throughout the paper ,
we construct exchangeable and consistent markov processes using a special _ rewiring measure _
( section [ section : exchangeable rewiring maps ] ) . in continuous - time
, markov processes can admit infinitely many jumps in arbitrarily small time intervals ; however , by the consistency assumption , any edge can change only finitely often in bounded intervals . in this case
, we can choose a @xmath19-finite rewiring measure to direct the process .
for @xmath14 , an ( undirected ) _ graph _ @xmath40 with vertex set @xmath15 $ ] can be represented by its symmetric _ adjacency matrix _
@xmath44 for which @xmath69 if @xmath40 has an edge between @xmath21 and @xmath23 , and @xmath70 otherwise . by convention
, we always assume @xmath45 for all @xmath71 .
we write @xmath72 to denote the finite collection of all graphs with vertex set @xmath15 $ ] . on @xmath73 , we define the following operation of _ rewiring_. let @xmath74 be an @xmath75 symmetric matrix with entries in @xmath76 and all diagonal entries @xmath77 . for convenience , we write each entry of @xmath78 as a pair @xmath79 , @xmath80 .
we define a map @xmath81 by @xmath82 , where @xmath83 more compactly , we may write @xmath84 .
we call @xmath78 a _ rewiring map _ and @xmath85 the _ rewiring _ of @xmath40 by @xmath78 .
we write @xmath86 to denote the collection of all rewiring maps @xmath87 , which are in one - to - one correspondence with @xmath75 symmetric matrices with entries in @xmath76 and all diagonal entries @xmath77 .
the following display illustrates the rewiring operation in ( [ eq : rewire ] ) . given @xmath88 and @xmath89
, we obtain @xmath85 by choosing the appropriate element of each entry of @xmath78 : if @xmath70 , we choose the _ left _ coordinate of @xmath90 ; if @xmath69 , we choose the _ right _ coordinate of @xmath90 .
for example , @xmath91 a unique symmetric @xmath75 matrix determines each element in @xmath72 and @xmath92 , and so there is a natural restriction operation on both spaces by taking the leading @xmath93 submatrix , for any @xmath94 .
in particular , we write @xmath95}:= ( g_{ij})_{1\leq i , j\leq m},\quad\quad g \in \mathcal{g}_n , \quad\mbox{and}\nonumber \\[-8pt]\\[-8pt ] w|_{[m]}&:=&(w_{ij})_{1\leq i ,
j\leq m},\quad\quad w\in \mathcal { w}_n , \nonumber\end{aligned}\ ] ] to denote the restrictions of @xmath96 and @xmath97 to @xmath98 and @xmath99 , respectively .
these restriction operations lead to the notions of _ infinite graphs _ and _ infinite rewiring maps _ as infinite symmetric arrays with entries in the appropriate space , either @xmath43 or @xmath100 .
we write @xmath101 to denote the space of infinite graphs , identified by a @xmath43-valued _ adjacency array _ , and @xmath102 to denote the space of infinite rewiring maps , identified by a symmetric @xmath76-valued array with @xmath77 on the diagonal .
any @xmath103 acts on @xmath101 just as in ( [ eq : rewire ] ) and , for any @xmath104 , the rewiring operation satisfies @xmath105}=w|_{[n]}(g|_{[n ] } ) \quad\quad\mbox{for all } n\in \mathbb{n}.\ ] ] the spaces @xmath101 and @xmath102 are uncountable but can be equipped with the discrete @xmath19-algebras @xmath106 and @xmath107 , respectively , so that the restriction maps @xmath108}$ ] are measurable for every @xmath109 . moreover , both @xmath101 and @xmath102 come equipped with a product - discrete topology induced , for example , by the ultrametric @xmath110}=w'|_{[n]}\bigr\},\quad\quad w , w'\in \wiren.\ ] ] the metric on @xmath101 is analogous .
both @xmath101 and @xmath102 are compact , complete , and separable metric spaces .
much of our development hinges on the following proposition , whose proof is straightforward .
[ prop : lipschitz ] rewiring maps are associative under composition and lipschitz continuous in the metric ( [ eq : metric ] ) , with lipschitz constant 1 .
let @xmath111 denote the collection of _ finite _ permutations of @xmath67 , that is , permutations @xmath112 for which @xmath113 .
we call any random array @xmath114 _ weakly exchangeable _ if @xmath115 and @xmath116 where @xmath117 denotes _ equality in law_. aldous defines weak exchangeability using only the latter condition ; see @xcite , chapter 14 , page 132 .
we impose symmetry for convenience in this paper , all graphs and rewiring maps are symmetric arrays . from the discussion in section [ section : exchangeability ] , we are interested in models for random graphs @xmath118 that are _ exchangeable _ , meaning the adjacency matrix @xmath119 is a weakly exchangeable @xmath43-valued array .
likewise , we call a random rewiring map @xmath120 _ exchangeable _ if its associated @xmath100-valued array @xmath121 is weakly exchangeable .
de finetti s theorem represents any infinitely exchangeable sequence @xmath122 in a polish space @xmath123 with a ( non - unique ) measurable function @xmath124 ^ 2\rightarrow\mathcal{s}$ ] such that @xmath125 , where @xmath126 for @xmath127 independent , identically distributed ( i.i.d . ) uniform random variables on @xmath128 $ ] .
the aldous
hoover theorem @xcite extends de finetti s representation ( [ eq : de finetti rep ] ) to weakly exchangeable @xmath123-valued arrays : to any such array @xmath129 , there exists a ( non - unique ) measurable function @xmath130 ^ 4\rightarrow\mathcal{s}$ ] satisfying @xmath131 such that @xmath132 , where @xmath133 for @xmath134 i.i.d .
uniform random variables on @xmath128 $ ] .
the function @xmath135 has a statistical interpretation that reflects the structure of the random array .
in particular , @xmath135 decomposes the law of @xmath136 into individual @xmath137 , row @xmath138 , column @xmath139 , and overall @xmath140 effects .
the overall effect plays the role of the mixing measure in the de finetti interpretation .
if @xmath141 in ( [ eq : de finetti rep ] ) is constant with respect to its first argument , that is , @xmath142 for all @xmath143 $ ] , then @xmath144 constructed in ( [ eq : de finetti rep ] ) is an i.i.d . sequence .
letting @xmath141 vary with its first argument produces a mixture of i.i.d .
a fundamental interpretation of de finetti s theorem is : @xmath145 similarly , if @xmath135 in ( [ eq : a - h rep ] ) satisfies @xmath146 for all @xmath147 $ ] , then @xmath148 is _ dissociated _ , that is @xmath149}\mbox { is independent of } x^*|_{\{n+1,n+2,\ldots\ } } \mbox { for all } n\in\mathbb{n}.\ ] ] the aldous
hoover representation ( [ eq : a - h rep ] ) spurs the sequel to de finetti s interpretation : @xmath150 see aldous @xcite , chapter 14 , for more details .
we revisit the theory of weakly exchangeable arrays in section [ section : exchangeable rewiring maps ] .
throughout the paper , we use the rewiring maps to construct markov chains on @xmath101 . from any probability distribution @xmath151 on @xmath152
, we generate @xmath153 i.i.d . from @xmath151 and a random graph @xmath154 ( independently of @xmath153 ) .
we then define a markov chain @xmath155 on @xmath72 by @xmath156 we call @xmath151 _ exchangeable _ if @xmath157 is an exchangeable rewiring map , that is , @xmath158 for all permutations @xmath159\rightarrow[n]$ ] .
[ prop : exchangeable rewiring chain ] let @xmath151 be an exchangeable probability measure on @xmath152 and let @xmath160 be as constructed in ( [ eq : rewiring chain ] ) from an exchangeable initial state @xmath161 and @xmath153 i.i.d . from @xmath151 .
then @xmath11 is an exchangeable markov chain on @xmath72 with transition probability @xmath162 the markov property is immediate by mutual independence of @xmath163 .
the formula for the transition probabilities ( [ eq : rewiring tps ] ) follows by description ( [ eq : rewiring chain ] ) of @xmath11 .
we need only show that @xmath11 is exchangeable . by assumption
, @xmath161 is an exchangeable random graph on @xmath16 vertices , and so its distribution is invariant under arbitrary permutation of @xmath15 $ ] .
moreover , the law of @xmath157 satisfies @xmath158 and , for any fixed @xmath164 and @xmath165 , @xmath166 satisfies @xmath167 the @xmath33-entry of @xmath168 . therefore , @xmath169 and , for any exchangeable graph @xmath118 and exchangeable rewiring map @xmath120 , we have @xmath170 hence , the transition law of @xmath11 is equivariant with respect to relabeling .
since the initial state @xmath161 is exchangeable , so is the markov chain .
we call @xmath155 an _ @xmath171-rewiring markov chain_. from the discussion in section [ section : rewiring maps ] , we can define an exchangeable measure @xmath172 on @xmath92 as the restriction to @xmath92 of an exchangeable probability measure @xmath173 on @xmath102 , where @xmath174}=w\bigr\}\bigr),\quad\quad w\in \mathcal { w}_n .\ ] ] denote by @xmath175 the transition probability measure of an @xmath172-rewiring markov chain on @xmath176 , as defined in ( [ eq : rewiring tps ] ) .
[ thm : consistent rewiring chain ] for any exchangeable probability measure @xmath173 on @xmath102 , @xmath177 is a consistent family of exchangeable transition probabilities in the sense that @xmath178 for every @xmath179}=g\}$ ] , where @xmath180 is defined in ( [ eq : restriction - graph ] ) .
proposition [ prop : exchangeable rewiring chain ] implies exchangeability of @xmath181 .
it remains to show that @xmath181 satisfies ( [ eq : consistent tps ] ) . by ( [ eq : rewiring tps ] ) , @xmath182 now , for any @xmath183 , fix @xmath184 and @xmath185 .
then ( [ eq : consistent tps ] ) requires @xmath186 which follows by definition ( [ eq : omega - n omega ] ) of @xmath187 . to see this , note that @xmath188}(g)=g'\bigr\}\bigr)&=&\omega\bigl(\bigl\{w\in \wiren\dvt ( w|_{[n]})|_{[m]}(g)=g'\bigr\}\bigr ) \\ & = & \omega\bigl(\bigl\{w\in\wiren\dvt w|_{[m]}(g)=g'\bigr\ } \bigr ) \\ & = & \omega^{(m)}\bigl(\bigl\{w\in\wirem\dvt w(g)=g'\bigr\ } \bigr).\end{aligned}\ ] ] this completes the proof .
the consistency condition ( [ eq : consistent tps ] ) for markov chains is exactly the necessary and sufficient condition for a function of a markov chain to be a markov chain , as proven in @xcite . before describing the measure @xmath173 from theorem [ thm : consistent rewiring chain ] in further detail , we first show some concrete examples of rewiring chains . for any @xmath189 ,
let @xmath190 denote the erds r ' enyi measure on @xmath101 , which we define by its finite - dimensional restrictions @xmath191 to @xmath176 for each @xmath30 , @xmath192 given any pair @xmath193\times[0,1]$ ] , the _ @xmath194-erds
rnyi chain _ has finite - dimensional transition probabilities @xmath195 for @xmath196 , the @xmath194-erds rnyi rewiring chain has unique stationary distribution @xmath197 , with @xmath198 . by assumption , both @xmath42 and @xmath41 are strictly between @xmath54 and @xmath55 and , thus , ( [ eq : er fidi ] ) assigns positive probability to every transition in @xmath73 , for every @xmath199 .
therefore , each finite - dimensional chain is aperiodic and irreducible , and each possesses a unique stationary distribution @xmath200 . by consistency of the transition probabilities @xmath201 ( theorem [ thm : consistent rewiring chain ] ) , the finite - dimensional stationary measures @xmath202 must be exchangeable and consistent and , therefore , they determine a unique measure @xmath203 on @xmath101 , which is stationary for @xmath204 .
furthermore , by conditional independence of the edges of @xmath39 , given @xmath40 , the stationary law must be erds rnyi with some parameter @xmath205 . in an @xmath206-random graph ,
all edges are present or not independently with probability @xmath207 .
therefore , it suffices to look at the probability of the edge between vertices labeled 1 and 2 . in this case , we need to choose @xmath207 so that @xmath208 which implies @xmath209 .
some elementary special cases of the @xmath194-erds rnyi rewiring chain are worth noting .
first , for @xmath194 either @xmath77 or @xmath210 , this chain is degenerate at either the empty graph @xmath211 or the complete graph @xmath212 and has unique stationary measure @xmath213 or @xmath214 , respectively . on the other hand ,
when @xmath215 , the chain is degenerate at its initial state and so its initial distribution is stationary .
however , if @xmath216 , then the chain alternates between its initial state @xmath40 and its complement @xmath217 , where @xmath218 for all @xmath219 ; in this case , the chain is periodic and does not have a unique stationary distribution .
we also note that when @xmath220 for some @xmath221 , the chain is simply an i.i.d .
sequence of @xmath190-random graphs with stationary distribution @xmath197 , where @xmath222 , as it must . for @xmath59
, we define the _ mixed erds rnyi rewiring chain _ through @xmath64 , the mixture of @xmath223-laws with respect to the beta law with parameter @xmath224 . writing @xmath225 we derive @xmath226 } \varepsilon _ p^{(n)}(g)\frac{\gamma(\alpha+\beta)}{\gamma(\alpha)\gamma ( \beta)}p^{\alpha-1}(1-p)^{\beta-1}\,\mathrm{d}p \\[-1pt ] & = & \frac{\gamma(\alpha+\beta)}{\gamma(\alpha)\gamma(\beta ) } \frac{\gamma(\alpha+n_1)\gamma(\beta+n_0)}{\gamma(\alpha+\beta + n)}\int_{[0,1 ] } \mathscr{b}_{\alpha+n_1,\beta+n_0}(\mathrm{d}p ) \\[-1pt ] & = & \frac{\alpha^{\uparrow n_1}\beta^{\uparrow n_0}}{(\alpha+\beta ) ^{\uparrow n}},\end{aligned}\ ] ] where @xmath227 , @xmath63 , and @xmath228 . for @xmath229
, we define _ mixed erds
rnyi transition probabilities _ by @xmath230 an interesting special case takes @xmath231 and @xmath232 for @xmath233 . in this case ,
( [ eq : mixture er ] ) becomes @xmath234 @xmath235 is reversible with respect to @xmath236 . for fixed @xmath237 ,
we write @xmath238 and @xmath239 .
note that @xmath240 .
therefore , we have @xmath241 & = & \frac{\alpha^{\uparrow n'_{00}}\beta^{\uparrow n'_{10}}\alpha ' ^{\uparrow n'_{11}}\beta^{\uparrow n'_{01}}}{(\alpha+2\beta+\alpha ' ) ^{\uparrow n } } \\[-1pt ] & = & \varepsilon^{(n)}_{\alpha+\beta,\alpha'+\beta}\bigl(g ' \bigr)p_{(\beta , \alpha),(\alpha',\beta)}^{(n)}\bigl(g',g\bigr),\end{aligned}\ ] ] establishing detailed balance and , thus , reversibility . a mixed erds rnyi markov chain is directed by @xmath242\times[0,1]}\omega_{p_0,p_1}(\mathrm{d}w ) ( \mathscr { b}_{\alpha_0,\beta_0}\otimes\mathscr{b}_{\alpha_1,\beta _ 1 } ) ( \mathrm{d}p_0,\mathrm{d}p_1),\quad\quad w\in\wiren,\ ] ] where @xmath243 is determined by its finite - dimensional distributions @xmath244 for @xmath196 , for every @xmath30 . in the next section
, we see that a representation of the directing measure @xmath173 as a mixture of simpler measures holds more generally .
notice that @xmath245 is dissociated for all fixed @xmath246 . by the aldous
hoover theorem , we can express any exchangeable measure on @xmath102 as a mixture of dissociated measures .
to more precisely describe the mixing measure @xmath173 , we extend the theory of _ graph limits _ to its natural analog for rewiring maps .
we first review the related theory of graph limits , as surveyed by lov ' asz @xcite .
a graph limit is a statistic that encodes a lot of structural information about an infinite graph .
in essence , the graph limit of an exchangeable random graph contains all relevant information about its distribution . for any injection @xmath247\rightarrow[n]$ ] , @xmath183 , and @xmath248 , we define @xmath249 . in words , @xmath250 is the subgraph @xmath40 induces on @xmath251 $ ] by the vertices in the range of @xmath252 .
given @xmath253 and @xmath254 , we define @xmath255 to equal the number of injections @xmath247\rightarrow [ n]$ ] such that @xmath256 . intuitively , @xmath255 is the number of `` copies '' of @xmath257 in @xmath40 , which we normalize to obtain the _ density _ of @xmath257 in @xmath40 , @xmath258 where @xmath259 is the number of unique injections @xmath247\rightarrow[n]$ ] . the _ limiting density _ of @xmath257 in any infinite graph @xmath104 is @xmath260}),\quad\quad f\in \graphsm , \quad\quad\mbox{if it exists}.\ ] ] the collection @xmath261 is countable and so we can define the _ graph limit _ of @xmath262 by @xmath263 provided @xmath264 exists for all @xmath265 .
any graph limit is an element in @xmath128^{\mathcal{g}^*}$ ] , which is compact under the metric @xmath266^{\mathcal{g}^*}.\ ] ] the space of graph limits is a compact subset of @xmath128^{\mathcal { g}^{*}}$ ] , which we denote by @xmath267 .
we implicitly equip @xmath128^{\mathcal{g}^*}$ ] with its borel @xmath19-field and @xmath267 with its trace @xmath19-field .
any @xmath268 is a sequence @xmath269 , where @xmath270 denotes the coordinate of @xmath271 corresponding to @xmath272 . in this way
, any @xmath268 determines a probability measure @xmath273 on @xmath73 , for every @xmath30 , by @xmath274 furthermore , the collection @xmath275 is consistent and exchangeable on @xmath276 and , by kolmogorov s extension theorem , determines a unique exchangeable measure @xmath277 on @xmath101 , for which @xmath277-almost every @xmath104 has @xmath278 .
conversely , combining the aldous
hoover theorem for weakly exchangeable arrays ( @xcite , theorem 14.21 ) and lovsz szegedy theorem of graph limits ( @xcite , theorem 2.7 ) , any exchangeable random graph @xmath118 is governed by a mixture of @xmath277 measures .
in particular , to any exchangeable random graph @xmath118 , there exists a unique probability measure @xmath279 on @xmath267 such that @xmath280 , where @xmath281 since @xmath76 is a finite space , the aldous
hoover theorem applies to exchangeable rewiring maps .
following section [ section : graph limits ] , we define the _ density _ of @xmath282 in @xmath283 by @xmath284 where @xmath285 equals the number of injections @xmath286\rightarrow[n]$ ] for which @xmath287 . for an infinite rewiring map @xmath288
, we define @xmath289}),\quad\quad w\in\mathcal { w}_m,\quad\quad \mbox{if it exists}.\ ] ] as for graphs , the collection @xmath290 is countable and so we can define the _ rewiring limit _ of @xmath288 by @xmath291 provided @xmath292 exists for all @xmath293 . we write @xmath294^{\mathcal{w}^*}$ ] to denote the compact space of rewiring limits and @xmath295 to denote the coordinate of @xmath296 corresponding to @xmath293 .
we equip @xmath297 with the metric @xmath298 [ lemma : rewiring structure ] every @xmath296 satisfies * @xmath299}=v\ } } \upsilon(v^*)$ ] for every @xmath300 , for all @xmath30 , and * @xmath301 for every @xmath30 . by definition of @xmath297
, we may assume that @xmath302 is the rewiring limit @xmath303 of some @xmath304 so that @xmath305 , for every @xmath293 .
from the definition of the rewiring limit ( [ eq : rewiring limit ] ) , @xmath306})}{n^{\downarrow m}}=\lim _ { n\rightarrow\infty}\sum_{w\in\wirem } \frac{\ind ( w , w^*|_{[n]})}{n^{\downarrow m}}=1,\ ] ] where the interchange of sum and limit is justified by the bounded convergence theorem because @xmath307})/n^{\downarrow m}\leq1 $ ] for all @xmath308 . also , for every @xmath183 and @xmath309 , we have @xmath310}=w\ } } \upsilon\bigl(w'\bigr)&= & \sum_{\{w'\in \mathcal{w}_n : w'|_{[m]}=w\}}\lim_{k\rightarrow\infty } \frac{\ind ( w',w^*|_{[k]})}{k^{\downarrow n } } \\ & = & \lim_{k\rightarrow\infty}\sum_{\{w'\in \mathcal { w}_n : w'|_{[m]}=w\ } } \frac{\ind ( w',w^*|_{[k]})}{k^{\downarrow n } } \\ & = & \lim_{k\rightarrow\infty}\frac{\ind ( w , w^*|_{[k]})}{k^{\downarrow m } } \\ & = & v(w).\end{aligned}\ ] ] this follows by the definition of @xmath311 and also because , for any @xmath247\rightarrow[k]$ ] there are @xmath312 injections @xmath313\rightarrow[k]$ ] such that @xmath314 coincides with @xmath252 on @xmath251 $ ] .
[ lemma : compact rewiring ] @xmath315 is a compact metric space .
[ prop : dissociated ] let @xmath120 be a dissociated exchangeable rewiring map .
then , with probability one , @xmath316 exists and is nonrandom .
we delay the proofs of lemma [ lemma : compact rewiring ] and theorem [ prop : dissociated ] until section [ section : proof ] .
[ cor : existence rewiring limit ] let @xmath288 be an exchangeable random rewiring map
. then @xmath316 exists almost surely . by theorem [ prop : dissociated ] , every dissociated rewiring map possesses a nonrandom rewiring limit almost surely . by the aldous
hoover theorem , @xmath120 is a mixture of dissociated rewiring maps and the conclusion follows . by lemma [ lemma : rewiring structure ] , any @xmath317 determines a probability measure @xmath318 on @xmath102 in a straightforward way : for each @xmath30 , we define @xmath319 as the probability distribution on @xmath152 with @xmath320 [ prop : rewiring measure ] for any @xmath296 , @xmath321 is a collection of exchangeable and consistent probability distributions on @xmath322 . in particular , @xmath321 determines a unique exchangeable probability measure @xmath323 on @xmath102 for which @xmath318-almost every @xmath103 has @xmath324 . by lemma [ lemma : rewiring structure ] ,
the collection @xmath325 in ( [ eq : measure on wiren ] ) is a consistent family of probability distributions on @xmath326 .
exchangeability follows because @xmath327})$ ] is invariant under relabeling of @xmath78 , that is , @xmath328}^{\sigma'})$ ] for all permutations @xmath329 and @xmath330 . by kolmogorov
s extension theorem , @xmath331 determines a unique measure @xmath323 on the limit space @xmath102 .
finally , @xmath332 is dissociated and so , by theorem [ prop : dissociated ] , @xmath333 almost surely .
we call @xmath323 in proposition [ prop : rewiring measure ] a _ rewiring measure _ directed by @xmath302 . for any measure @xmath334 on @xmath297
, we define the @xmath335-mixture of rewiring measures by @xmath336 to any exchangeable rewiring map @xmath120 , there exists a unique probability measure @xmath334 on @xmath297 such that @xmath337 .
this follows by the aldous hoover theorem and proposition [ prop : rewiring measure ] . from theorem [ prop : dissociated ] and proposition [ prop : rewiring measure ] , any probability measure @xmath334 on @xmath297 corresponds to an @xmath338-rewiring chain as in theorem [ thm : consistent rewiring chain ] .
we now refine our discussion to rewiring chains in continuous - time , for which infinitely many transitions can `` bunch up '' in arbitrarily small intervals , but individual edges jump only finitely often in bounded intervals . henceforth , we write @xmath339 to denote the identity @xmath340 and , for @xmath30 , we write @xmath341 to denote the identity @xmath342 .
let @xmath173 be an exchangeable measure on @xmath102 such that @xmath343}\neq\idn\}\bigr)<\infty \quad\quad\mbox{for every } n\geq2.\ ] ] similar to our definition of @xmath344 in section [ section : discrete ] , we use @xmath173 to define the transition rates of continuous - time @xmath173-rewiring chain .
briefly , we assume @xmath345 because the identity map @xmath346 is immaterial for continuous - time processes .
the finiteness assumption on the right of ( [ eq : regularity omega ] ) ensures that the paths of the finite restrictions are cdlg . for each @xmath30 , we write @xmath172 to denote the restriction of @xmath173 to @xmath152 and define @xmath347 [ prop : finite q - omega ] for each @xmath30
, @xmath348 is a finite , exchangeable conditional measure on @xmath349 .
moreover , the collection @xmath350 satisfies @xmath351 for all @xmath352 , for all @xmath183 , for every @xmath30 , where @xmath180 is the restriction map @xmath353 defined in ( [ eq : restriction - graph ] ) .
finiteness of @xmath348 follows from ( [ eq : regularity omega ] ) since , for every @xmath253 , @xmath354 exchangeability of @xmath348 follows by proposition [ prop : exchangeable rewiring chain ] and exchangeability of @xmath173 .
consistency of @xmath350 results from lipschitz continuity of rewiring maps ( proposition [ prop : lipschitz ] ) and consistency of the finite - dimensional marginals @xmath355 associated to @xmath173 : for fixed @xmath356 and @xmath352 , @xmath357}=g'}q_{\omega}^{(n ) } \bigl(g^*,g''\bigr ) \\ & = & \sum_{g'':g''|_{[m]}=g'}\omega^{(n)}\bigl(\bigl\{w\in \mathcal { w}_n \dvt w\bigl(g^*\bigr)=g''\bigr\ } \bigr ) \\ & = & \omega^{(n)}\bigl(\bigl\{w\in \mathcal{w}_n \dvt w|_{[m]}(g)=g'\bigr\}\bigr ) \\ & = & \omega^{(m)}\bigl(\bigl\{w\in\wirem\dvt w(g)=g'\bigr\ } \bigr ) \\ & = & q_{\omega}^{(m)}\bigl(g , g'\bigr).\end{aligned}\ ] ] from @xmath358 , we define a collection of infinitesimal jump rates @xmath359 by @xmath360 [ cor : inf - q ] the infinitesimal generators @xmath359 are exchangeable and consistent and , therefore , define the infinitesimal jump rates @xmath361 of an exchangeable markov process on @xmath101 .
consistency when @xmath362 was already shown in proposition [ prop : finite q - omega ] .
we must only show that @xmath363 is consistent for @xmath364 .
fix @xmath30 and @xmath253 . then , for any @xmath365 , we have @xmath366 in section [ section : informal description ] , we mentioned local and global discontinuities for graph - valued processes . in the next two sections ,
we formally incorporate these discontinuities into a continuous - time rewiring process : in section [ section : the rewiring measure ] , we extend the notion of random rewiring from discrete - time ; in section [ section : local - edge ] , we introduce transitions for which , at the time of a jump , only a single edge in the network changes . over time
, the local changes can accumulate to cause a non - trivial change to network structure . in this section
, we specialize to the case where @xmath367 for some measure @xmath334 on @xmath297 satisfying @xmath368 where @xmath369 is the rewiring limit of @xmath370 and @xmath371 is the entry of @xmath302 corresponding to @xmath341 , for each @xmath30 . for each @xmath199 , we write @xmath372 to denote @xmath373 for @xmath374 , and likewise for the infinitesimal generator @xmath375 .
[ lemma : upsilon - omega equiv ] for @xmath334 satisfying ( [ eq : regularity upsilon ] ) , the rewiring measure @xmath338 satisfies ( [ eq : regularity omega ] ) . by theorem [ prop : dissociated ] , @xmath376 implies @xmath377 .
we need only show that @xmath378 implies @xmath379 for every @xmath380 . for any @xmath296 , @xmath381}\neq\idn\}\bigr)&= & \omega _ { \upsilon } \biggl(\bigcup_{1\leq i < j\leq n}\{w\in \wiren\dvt w|_{\{i , j\ } } \neq\idij\ } \biggr ) \\
& \leq&\sum_{1\leq i < j\leq n}\omega_{\upsilon}\bigl(\{w\in \wiren\dvt w|_{\ { i , j\}}\neq\idij\}\bigr ) \\ & = & \sum_{1\leq i
< j\leq n}\omega_{\upsilon}^{(2 ) } \bigl(\mathcal { w}_2\setminus\{\idtwo\}\bigr ) \\ & = & \frac{n(n-1)}{2}\bigl(1-\upsilon^{(2)}_*\bigr).\end{aligned}\ ] ] hence , by ( [ eq : regularity upsilon ] ) , @xmath382}\neq\idn\}\bigr)\leq \int_{\rewiringlimits}\frac{n(n-1)}{2}\bigl(1-\upsilon^{(2 ) } _ * \bigr)\upsilon ( \mathrm{d}\upsilon)<\infty,\ ] ] for every @xmath380 . for each @xmath30
, @xmath372 is a finite , exchangeable conditional measure on @xmath349 .
moreover , @xmath383 satisfies @xmath384 this follows directly from propositions [ prop : rewiring measure ] , [ prop : finite q - omega ] , and lemma [ lemma : upsilon - omega equiv ] .
we may , therefore , define an infinitesimal generator for a markov chain on @xmath73 by @xmath385 [ thm : upsilon - rewiring process ] for each @xmath334 satisfying ( [ eq : regularity upsilon ] ) , there exists an exchangeable markov process @xmath11 on @xmath386 with finite - dimensional transition rates as in ( [ eq : infinitesimal - q ] ) .
we call @xmath11 in theorem [ thm : upsilon - rewiring process ] a _ rewiring process _ directed by @xmath334 , or with rewiring measure @xmath338 . for @xmath387 and @xmath388 ,
let @xmath389 denote the rewiring map @xmath390 that acts by mapping @xmath391 , @xmath392 in words , @xmath393 puts an edge between @xmath21 and @xmath23 ( if @xmath394 ) or no edge between @xmath21 and @xmath23 ( if @xmath395 ) and keeps every other edge fixed . for fixed @xmath30 , let @xmath396 denote the _ empty graph
, that is , the graph with no edges .
we generate a continuous - time process @xmath397 on @xmath349 as follows .
first , we specify a constant @xmath398 and , independently for each pair @xmath399\times[n]$ ] , @xmath34 , we generate i.i.d .
random variables @xmath400 from the exponential distribution with rate parameter @xmath401 . given @xmath402
, we define @xmath403 by @xmath404 where @xmath405 denotes an edge between @xmath21 and @xmath23 in @xmath40 .
clearly , @xmath403 is exchangeable and converges to a unique stationary distribution @xmath406 , the point mass at the complete graph @xmath407 .
moreover , the distribution of @xmath408 , the time until absorption in @xmath407 , is simply the law of the maximum of @xmath409 i.i.d . exponential random variables with rate parameter @xmath401 .
conversely , we could consider starting in @xmath407 , the _ complete graph _ , and generating the above process in reverse .
in this case , we specify @xmath410 and let @xmath411 be an i.i.d . collection of exponential random variables with rate parameter @xmath412 .
we construct @xmath413 , given @xmath411 , by @xmath414 for @xmath415 , this process evolves exactly as the _ complement _ of @xmath403 , that is , @xmath416 where @xmath417 is defined by @xmath418 for all @xmath419 and all @xmath1 .
it is natural to consider the superposition of @xmath403 and @xmath420 , which we call a _
@xmath421-local - edge process_. let @xmath422 and let @xmath423 denote the point mass at the single - edge update map @xmath424 .
following section [ section : exch rw ] , we define @xmath425 for @xmath422 , @xmath426 defined in ( [ eq : omega - c0c1 ] ) satisfies ( [ eq : regularity omega ] ) .
since @xmath427 only charges the single - edge update maps , it is clear that it assigns zero mass to the identity map . also , for any @xmath30 , the restriction of @xmath424 to @xmath152 coincides with the identity @xmath428 except when @xmath429 .
hence , @xmath430 for every @xmath380 . for any @xmath422
, there exists an exchangeable markov process on @xmath101 with jump rates given by @xmath427 . for every @xmath30 , the total jump rate out of any @xmath248 can be no larger than @xmath431 and so the finite - dimensional hold times are almost surely positive and the process on @xmath73 has cdlg sample paths .
the markov property and exchangeability follow by independence of the exponential hold times @xmath432 and corollary [ cor : inf - q ] .
consistency is apparent by the construction from independent poisson point processes .
this completes the proof . for any measure @xmath334 satisfying ( [ eq : regularity upsilon ] ) , @xmath422
, we call a rewiring process with jump measure @xmath433 an _ @xmath434-rewiring process_. from our discussion in this section , the @xmath435-rewiring process exists for any choice of @xmath334 satisfying ( [ eq : regularity upsilon ] ) and @xmath436 .
individually , @xmath338 and @xmath427 satisfy ( [ eq : regularity omega ] ) and , thus , so does @xmath437 .
furthermore , the family of @xmath438-rewiring processes is markovian , exchangeable , and consistent .
we can construct an @xmath439-rewiring process from a poisson point process . for @xmath440 , where @xmath334 satisfies ( [ eq : regularity upsilon ] ) and @xmath422 , let @xmath441 be a poisson point process with intensity @xmath442 . to begin , we take @xmath161 to be an exchangeable random graph and , for each @xmath30 , we define @xmath31}:=(\gamma^{[n]}_t)_{t\geq0}$ ] on @xmath73 by @xmath13}_0:=\gamma_0|_{[n]}$ ] and * if @xmath18 is an atom time of @xmath443 such that @xmath444}_t:=w_t|_{[n]}\neq\idn$ ] , then we put @xmath445}_{t}:=w^{[n]}_t(\gamma^{[n]}_{t-})$ ] ; * otherwise , we put @xmath13}_t=\gamma^{[n]}_{t-}$ ] . [
prop : thinned ppp ] for each @xmath30 , @xmath31}$ ] is a markov chain on @xmath73 with infinitesimal jump rates @xmath363 in ( [ eq : infinitesimal - omega ] ) .
we can define @xmath446}:=\{(t , w_t^{[n]})\}\subset\mathbb { r}^{+}\times \mathcal{w}_n $ ] from @xmath443 by removing any atom times for which @xmath447}:=w_t|_{[n]}=\idn$ ] , and otherwise putting @xmath447}:=w_t|_{[n]}$ ] . by the thinning property of poisson point processes , @xmath446}$ ]
is a poisson point process with intensity @xmath448 , where @xmath449 given @xmath450}=g$ ] , the jump rate to state @xmath362 is @xmath451 and the conclusion follows .
[ thm : existence rewiring ] for any @xmath173 satisfying ( [ eq : regularity omega ] ) , the @xmath452-rewiring process on @xmath101 exists and can be constructed from a poisson point process with intensity @xmath442 as above .
let @xmath443 be a poisson point process with intensity @xmath453 and construct @xmath454}\}_{n\in\mathbb{n}}$ ] from the thinned processes @xmath455}\}_{n\in\mathbb{n}}$ ] determined by @xmath443 . by proposition [ prop : thinned ppp
] , each @xmath31}$ ] is an exchangeable markov chain governed by @xmath363 .
moreover , @xmath454}\}_{n\in \mathbb{n}}$ ] is compatible by construction , that is , @xmath456}_t=\rmn\gamma^{[n]}_t$ ] for all @xmath1 , for all @xmath183 ; hence , @xmath454}\}_{n\in\mathbb{n}}$ ] defines a process @xmath11 on @xmath101 .
as we have shown previously , the infinitesimal rates given by @xmath457 are consistent and exchangeable ; hence , @xmath458 has infinitestimal generator @xmath361 and is an @xmath452-rewiring process .
any markov process @xmath11 on @xmath101 is characterized by its semigroup @xmath459 , defined as an operator on the space of continuous , bounded functions @xmath460 by @xmath461 where @xmath462 denotes the expectation operator with respect to the initial distribution @xmath463 , the point mass at @xmath40 .
we say @xmath11 has the _ feller property _ if , for all bounded , continuous functions @xmath464 , its semigroup satisfies * @xmath465 as @xmath466 for all @xmath104 , and * @xmath467 is continuous for all @xmath1 . the semigroup @xmath468 of any @xmath452-rewiring process enjoys the feller property . to show the first point in the feller property , we let @xmath104 and @xmath469 be an @xmath173-rewiring process with initial state @xmath470 and directing measure @xmath452 satisfying ( [ eq : regularity omega ] ) .
we define @xmath471}=g'|_{[n ] } \rightarrow h(g)=h\bigl(g'\bigr)\bigr\}.\ ] ] by ( [ eq : regularity omega ] ) and finiteness of @xmath73 , @xmath13}_t\rightarrow g|_{[n]}$ ] in probability as @xmath466 , for every @xmath30 .
thus , for any @xmath472 , let @xmath473 be such that @xmath474 then @xmath475}\rightarrow g|_{[n]}$ ] in probability as @xmath466 and , therefore , @xmath465 by the bounded convergence theorem .
right - continuity at zero for all bounded , continuous @xmath464 follows by the stone weierstrass theorem .
for the second point , let @xmath476 have @xmath477 for some @xmath30 and construct @xmath11 and @xmath478 from the same poisson point process @xmath443 but with initial states @xmath470 and @xmath479 . by lipschitz continuity of the rewiring maps ( proposition [ prop : lipschitz ] ) , @xmath11 and @xmath478 can never be more than distance @xmath480 apart , for all @xmath1 .
continuity of @xmath481 , for each @xmath1 , follows . by the feller property
, any @xmath173-rewiring process has a cdlg version and its jumps are characterized by an infinitesimal generator . in section [ section
: continuous ] , we described the infinitesimal generator through its finite restrictions .
ethier and kurtz @xcite give an extensive treatment of the general theory of feller processes .
we have presented a family of time - varying network models that is markovian , exchangeable , and consistent , natural statistical properties that impose structure without introducing logical pitfalls .
external to statistics , exchangeable models are flawed : they produce _ dense _ graphs when conventional wisdom suggests real - world networks are _ sparse_. the erds rnyi model s storied history cautions against dismay . though it replicates little real - world network structure , the erds rnyi model has produced a deluge of insight for graph - theoretic structures and is a paragon of the utility of the probabilistic method @xcite . while our discussion is specific to exchangeable processes , the general descriptions in sections [ section : discrete ] and [ section : continuous ] can be used to construct processes that are not exchangeable , and possibly even sparse .
the most immediate impact of the rewiring process may be for analyzing information spread on dynamic networks . under the heading of _ finite markov information exchange _
( fmie ) processes , aldous @xcite recently surveyed interacting particle systems models for social network dynamics .
informally , fmie processes model a random spread of information on a network .
some of the easiest to describe fmie processes coincide with well - known interacting particle systems , such as the voter and contact processes ; others mimic certain social behaviors , for example , _ fashionista _ and _ compulsive gambler_. simulation is a valuable practical tool for developing intuition about intractable problems .
aldous s expository account contains some hard open problems for time - invariant networks .
considering these same questions on dynamic networks seems an even greater challenge . despite these barriers ,
policymakers and scientists alike desire to understand how trends , epidemics , and other information spread on networks .
the poisson point process construction in section [ section : poissonian structure ] could be fruitful for deriving practical answers to these problems .
in this section , we prove some technical results from our previous discussion .
we now show that @xmath315 is a compact metric space . recall that @xmath297 is equipped with the metric @xmath482 since @xmath128^{\mathcal{w}^*}$ ] is compact in this metric , it suffices to show that @xmath297 is a closed subset of @xmath128^{\mathcal{w}^*}$ ] . by lemma [ lemma : rewiring structure ]
, every @xmath317 satisfies @xmath483}=w}\upsilon \bigl(w^*\bigr ) \quad\quad\mbox{for every } w\in \mathcal{w}_n\ ] ] and @xmath484 for all @xmath30 .
then , for any @xmath485^{\mathcal{w}^*}\setminus\rewiringlimits$ ] , there must be some @xmath473 for which @xmath486}=w}x^{(n+1 ) } \bigl(w^*\bigr)\biggr{\vert}>0.\ ] ] for any @xmath487 , let @xmath488^{\mathcal { w}^*}\dvt \rho(x , x')<\delta\}$ ] denote the @xmath489-ball around @xmath490 .
now , take any @xmath491 . by this assumption , @xmath492 and so @xmath493 whence , @xmath494 by the triangle inequality
, we have @xmath495}}x^{(n+1 ) } \bigl(w^*\bigr)\biggr{\vert}\\ & \leq&\sum_{w\in\mathcal{w}_n}\bigl{\vert}x^{(n)}(w)-x'^{(n)}(w ) \bigr{\vert}+\sum_{w\in\mathcal{w}_n}\biggl{\vert}\sum _ { w^*:w^*|_{[n]}=w}\bigl(x^{(n+1)}\bigl(w^*\bigr)-x'^{(n+1 ) } \bigl(w^*\bigr)\bigr)\biggr{\vert}\\ & & { } + \sum_{w\in\mathcal{w}_n}\biggl{\vert}x'^{(n)}(w)- \sum_{w^*:w^*|_{[n]}=w}x'^{(n+1)}\bigl(w^ * \bigr)\biggr{\vert}\\ & \leq&\varepsilon_x/4+\sum_{w\in\mathcal{w}_n}\sum _ { w^*:w^*|_{[n]}=w}\bigl{\vert}x^{(n+1)}\bigl(w^ * \bigr)-x'^{(n+1)}\bigl(w^*\bigr)\bigr{\vert}\\ & & { } + \sum _ { w\in\mathcal{w}_n}\biggl{\vert}x'^{(n)}(w)-\sum _ { w^*:w^*|_{[n]}=w}x'^{(n+1)}\bigl(w^*\bigr ) \biggr{\vert}\\ & \leq&\varepsilon_x/4+\varepsilon_x/2+\sum _ { w\in\mathcal { w}_n}\biggl{\vert}x'^{(n)}(w)-\sum _ { w^*:w^*|_{[n]}=w}x'^{(n+1)}\bigl(w^*\bigr ) \biggr{\vert}.\end{aligned}\ ] ] therefore , @xmath496}=w}x'^{(n+1)}\bigl(w^ * \bigr)\biggr{\vert}\geq\varepsilon_x/4>0,\ ] ] which implies @xmath497^{\mathcal{w}^*}\setminus\rewiringlimits$ ] , meaning @xmath128^{\mathcal{w}^*}\setminus\rewiringlimits$ ] is open and @xmath297 is closed .
since @xmath128^{\mathcal{w}^*}$ ] is compact , so is @xmath297 .
this completes the proof .
assume that @xmath120 is an exchangeable and dissociated rewiring map .
by the aldous hoover theorem , we can assume @xmath120 is constructed from a measurable function @xmath130 ^ 4\rightarrow\{0,1\}\times\{0,1\}$ ] for which ( i ) @xmath498 and ( ii ) @xmath499 .
more precisely , we assume @xmath500 , for each @xmath219 , where @xmath501 are i.i.d .
uniform random variables on @xmath128 $ ] .
conditional on @xmath502 , we define the quantity @xmath503}=v\mid\alpha = a\},\quad\quad v\in\wirem , m\in \mathbb{n},\ ] ] which , by the fact that @xmath120 is dissociated , is independent of @xmath504 ; hence , we define the non - random quantity @xmath505}=v\}\mid\alpha\bigr)=p \{w|_{[m]}=v\}.\ ] ] recall , from section [ section : rewiring limits ] , the definition @xmath506}):=\frac{\ind(v , w|_{[n]})}{n^{\downarrow m}}:=\frac { 1}{n^{\downarrow m}}\sum _ { \mathrm{injections } \psi:[m]\rightarrow [ n]}\mathbf{1}\bigl\{w|_{[n]}^{\psi}=v\bigr \},\quad\quad n\in\mathbb{n}.\ ] ] for every @xmath507 , we also define @xmath508\rightarrow[n]}e \bigl(\mathbf{1}\bigl\{w|_{[n]}^{\psi}=v\bigr\ } \mid w|_{[k]}\bigr),\quad\quad k=0,1,\ldots , n.\ ] ] in particular , for every @xmath30 , we have @xmath509\rightarrow[n]}e \bigl(\mathbf{1}\bigl\{w|_{[n]}^{\psi}=v\bigr\ } \mid w|_{[0]}\bigr)=t(v , w)\ ] ] and @xmath510\rightarrow[n]}e \bigl(\mathbf{1}\bigl\{w|^{\psi}_{[n]}=v\bigr\ } \mid w|_{[n ] } \bigr)=t(v , w|_{[n]}).\ ] ] we wish to show that @xmath511})\rightarrow t(v , w)$ ] almost surely , for every @xmath282 , @xmath512 . to do this , we first show that @xmath513 is a martingale with respect to its natural filtration , for every @xmath30 .
we can then appeal to azuma s inequality and the borel
cantelli lemma to show that @xmath514 as @xmath515 .
note that @xmath516\rightarrow[n]}\sum _ { w\in \mathcal{w}_n } e\bigl(\mathbf{1}\{w|_{[n]}=w\ } \mid { w}|_{[k]}\bigr)\mathbf{1}\bigl\{w^{\psi}=v\bigr\}\ ] ] and @xmath517 } ) \mid m_{k , n}\bigr).\ ] ] on the inside , we have @xmath518 } ) \\ & & \quad = e \biggl ( \frac{1}{n^{\downarrow m}}\sum_{\mathrm{injections\ } \psi:[m]\rightarrow[n]}\sum _ { w\in \mathcal{w}_n } \mathbf{1}\bigl\{w^{\psi}=v\bigr\}e\bigl(\mathbf { 1}\bigl\{w|_{[n]}^{\psi}=w\bigr\}\mid w|_{[k]}\bigr ) \bigm| m_{k , n},w|_{[k ] } \biggr ) \\ & & \quad=\frac{1}{n^{\downarrow m}}\sum_{\mathrm{injections\ } \psi : [ m]\rightarrow[n]}\sum _ { w\in \mathcal{w}_n } \mathbf{1}\bigl\{w^{\psi}=v\bigr\}e\bigl ( \mathbf{1}\bigl\{w|_{[n]}^{\psi}=w\bigr\ } \mid w|_{[k ] } \bigr);\end{aligned}\ ] ] whence , @xmath519})\mid m_{k , n}\bigr ) \\ & & \quad = e \biggl(\frac { 1}{n^{\downarrow m}}\sum _ { \mathrm{injections\ } \psi:[m]\rightarrow [ n]}\sum_{w\in \mathcal{w}_n } \mathbf{1}\bigl\ { w^{\psi}=v\bigr\}e\bigl(\mathbf{1}\bigl\{w|_{[n]}^{\psi}=w \bigr\}\mid w|_{[k]}\bigr ) \bigm| m_{k , n } \biggr ) \\ & & \quad=\frac{1}{n^{\downarrow m}}\sum_{\mathrm{injections\ } \psi :
[ m]\rightarrow[n]}\sum _ { w\in \mathcal{w}_n } \mathbf{1}\bigl\{w^{\psi}=v\bigr\}e \bigl(e \bigl(\mathbf{1}\{w|_{[n]}=w\}\mid w|_{[k]}\bigr ) \mid m_{k , n } \bigr ) \\ & & \quad=\frac{1}{n^{\downarrow m}}\sum_{\mathrm{injections\ } \psi : [ m]\rightarrow[n]}\sum _ { w\in \mathcal{w}_n } \mathbf{1}\bigl\{w^{\psi}=v\bigr\}e\bigl ( \mathbf{1}\{w|_{[n]}=w\ } \mid m_{k , n}\bigr ) \\ & & \quad = m_{k , n}.\end{aligned}\ ] ] therefore , @xmath520 is a martingale for every @xmath30 .
furthermore , for every @xmath521 , @xmath522\rightarrow[n]}e\bigl(\mathbf{1}\bigl\{w|_{[n]}^{\psi } = v \bigr\}\mid w|_{[k+1]}\bigr)-e\bigl(\mathbf{1}\bigl\{w|_{[n]}^{\psi}=v \bigr\}\mid w|_{[k]}\bigr)\biggr{\vert}\\ & & \quad\leq\frac{1}{n^{\downarrow m}}\sum_{\mathrm{injections\ } \psi : [ m]\rightarrow[n]}\bigl|e\bigl ( \mathbf{1}\bigl\{w|_{[n]}^{\psi}=v\bigr\}\mid w|_{[k+1 ] } \bigr)-e\bigl(\mathbf{1}\bigl\{w|_{[n]}^{\psi}=v\bigr\}\mid w|_{[k]}\bigr)\bigr| \\ & & \quad\leq m(n-1)^{\downarrow(m-1)}/n^{\downarrow m } \\ & & \quad\leq m / n,\end{aligned}\ ] ] since @xmath523}^{\psi}=v\}\mid w|_{[k+1]})-e(\mathbf { 1}\{w|_{[n]}^{\psi}=v\}\mid w|_{[k]})=0 $ ] whenever @xmath252 does not map an element to @xmath524 .
the conditions for azuma s martingale inequality are thus satisfied and we have , for every @xmath525 , @xmath526 thus , @xmath527 and we conclude , by the borel
cantelli lemma , that @xmath528})-t(v , w)\bigr| > \varepsilon \bigr\ } \\ & & \quad= \bigl\ { \bigl|t(v , w|_{[n]})-t(v , w)\bigr|>\varepsilon\mbox { for infinitely many } n\in \mathbb{n } \bigr\}\end{aligned}\ ] ] has probability zero .
it follows that @xmath529})=t(v , w)$ ] exists with probability one for every @xmath530 .
therefore , with probability one , the rewiring limit @xmath531 exists .
we have already shown , by the assumption that @xmath120 is dissociated , that @xmath292 is non - random for every @xmath532 ; hence , the limit @xmath531 is non - random .
this completes the proof .
this work is partially supported by nsf grant dms-1308899 and nsa grant h98230 - 13 - 1 - 0299 . | we introduce the _ exchangeable rewiring process _ for modeling time - varying networks .
the process fulfills fundamental mathematical and statistical properties and can be easily constructed from the novel operation of _ random rewiring_. we derive basic properties of the model , including consistency under subsampling , exchangeability , and the feller property .
a reversible sub - family related to the erds rnyi model arises as a special case .
./style / arxiv - general.cfg |
functional equations play a very important role in the theory of zeta and @xmath0-functions . in the case of the most classical riemann zeta - function @xmath1 ,
riemann proved the beautiful symmetric functional equation @xmath2 this symmetry , however , does not remain when we consider the hurwitz zeta - function @xmath3 , where @xmath4 .
the functional equation for @xmath5 is of the form @xmath6 where @xmath7 and @xmath8 is the lerch zeta - function ( see ( * ? ? ? * ( 2.17.3 ) ) ) .
when @xmath9 , reduces to , but the form is no longer symmetric . in recent decades , the theory of various multiple zeta - functions has been developed extensively .
therefore it is natural to search for functional equations for those multiple zeta - functions .
for example , barnes multiple zeta - functions @xmath10 ( barnes @xcite ) with complex parameters @xmath11 is a direct generalization of hurwitz zeta - functions , so a kind of functional equation similar to is expected to hold .
in fact , such an equation has been proved under certain condition of parameters ( see hardy and littlewood @xcite in the case @xmath12 , and @xcite @xcite in general ) .
another important class of multiple zeta - functions is the multi - variable sum @xmath13 which was first considered by euler in the case @xmath12 , and then introduced by hoffman @xcite and zagier @xcite independently of each other for general @xmath14 . to find
some kind of functional equations for the sum or its variants / generalizations seems a rather complicated problem .
let us consider the simplest case @xmath12 : @xmath15 in the following sections we will report our attempt to search for functional equations for and its various variants .
let @xmath4 , @xmath16 , @xmath17 , and define @xmath18 this is a generalization of of the hurwitz - lerch type .
the functional equation for @xmath19 can be stated in terms of the confluent hypergeometric function @xmath20 where @xmath21 , @xmath22 , @xmath23 , @xmath24 .
define @xmath25 where @xmath26 then the following functional equation is known . [ thm1 ] the functions @xmath27
can be continued meromorphically to the whole space @xmath28 , and the functional equation @xmath29 holds . here
, the first term on the right - hand side of is an `` additional '' term , and the main body of the right - hand side is the second term involving @xmath30 .
this term , compared with the left - hand side , expresses a duality between the values at @xmath31 and at @xmath32 ( and also a duality between @xmath33 and @xmath34 ) , and hence formula can be regarded as a double analogue of .
the functions @xmath30 are not dirichlet series in the usual sense , but the confluent hypergeometric function satisfies the asymptotic expansion @xmath35 so we may say that @xmath36 can be approximated by @xmath37 . from this viewpoint it is possible to say that @xmath30 can be approximated by the dirichlet series @xmath38 in this sense @xmath30 may be regarded as generalized dirichlet series . the meromorphic continuation of @xmath30 was shown in ( * ? ? ?
* proposition 2 ) , where their functional equation @xmath39 was also proved .
the transformation formula @xmath40 of the confluent hypergeometric function is used in the proof of . applying to (
* proposition 1 ) , we can immediately obtain formula .
therefore the above theorem [ thm1 ] is essentially included in @xcite , though it is first explicitly stated in @xcite .
the main statement of @xcite is as follows .
let @xmath41 then [ thm2 ] we have @xmath42 and especially , when @xmath43 , we have @xmath44 this theorem can also be easily deduced from proposition 1 and proposition 2 of @xcite
. [ rem1 ] ( historical note ) the idea of the proof of theorem [ thm1 ] goes back to @xcite , where certain mean values of dirichlet @xmath0-functions were studied .
the application of the confluent hypergeometric function in this context was first done by katsurada @xcite . in order to study barnes double zeta - functions ( the case @xmath12 of )
, the second - named author @xcite introduced the two - variable double series @xmath45 ( the case @xmath46 of ) , and studied its properties , invoking the methods in @xcite , @xcite .
in particular , the special case @xmath46 of ( * ? ? ?
* proposition 1 ) was already given in ( * ? ? ?
* ( 5.5 ) ) .
in this and the next section we present the contents of two joint papers of komori , tsumura and the second - named author .
theorem [ thm1 ] is a non - symmetric functional equation , similarly to for hurwitz zeta - functions .
is there any _ symmetric _ functional equation for double zeta - functions ?
one of the main point of @xcite is that such equations do exist on certain special hyperplanes . in @xcite , the following generalization of
was introduced : @xmath47 where @xmath48 with @xmath49 , @xmath50 .
the functional equation similar to theorem [ thm1 ] holds for @xmath51 , which is @xmath52 ( ( * ? ? ?
* theorem 2.1 ) ) .
this formula itself can be proved just similarly to theorem [ thm1 ] . however , from this formula , it is possible to deduce the following symmetric functional equation .
let @xmath53 then we have [ thm3 ] the hyperplane @xmath54 @xmath55 is not a singular locus of @xmath51 , and when @xmath56 , the functional equation @xmath57 holds . when @xmath58 and @xmath59 , this theorem is actually almost immediately obtained from .
( but the second - named author did not notice this point when he wrote @xcite ) . from theorem [ thm3 ] , we can evaluate certain values of @xmath51 when @xmath60 are non - positive integers ( ( * ? ? ? * corollary 2.4 ) ) .
in the statement of aforementioned functional equations ( , ) , an `` additional '' term involving @xmath61 or @xmath62 appears as the first term of the right - hand side , which also appears in the definitions and .
the reason of the appearance of such a term can be clarified when we consider a little more general situation .
let @xmath63 , and let @xmath64 ( @xmath65 ) be periodic functions with period @xmath66 .
assume @xmath67 is an even ( or odd ) function , and let @xmath68 ( resp.@xmath69 ) if @xmath67 is even ( resp .
define the associated @xmath0-function by @xmath70 .
in @xcite the double series @xmath71 was introduced .
denote the finite fourier expansion of @xmath67 by @xmath72 the series @xmath73 can be continued meromorphically to the whole space @xmath28 . as an analogue of theorem [ thm3 ] , the following symmetric functional equation holds .
[ thm4 ] the functional equation @xmath74 holds on the hyperplane @xmath75 @xmath76 if @xmath77 , and on the hyperplane @xmath78 @xmath76 if @xmath79 . therefore if @xmath80 then the `` additional '' terms do not appear .
in particular , if @xmath81 are primitive dirichlet characters @xmath82 of conductor @xmath66 , then holds , and so the functional equation becomes the following very simple form : @xmath83 on the hyperplanes indicated in the statement of theorem [ thm4 ] .
[ rem2 ] using the relation @xmath84 ( where @xmath85 denotes the gauss sum ) , we can restate formula in which @xmath86 is replaced by @xmath87 .
such a formula is stated as ( * ? ? ?
* corollary 2.3 ) , but the factor @xmath88 is lacking on the left - hand side of the statement there . from this result
we can evaluate certain values of double @xmath0-functions at non - positive integer points ( ( * ? ? ?
* section 3 ) ) .
those results further motivates the construction of the theory of @xmath89-adic multiple @xmath0-functions .
a double analogue of the kubota - leopoldt @xmath89-adic @xmath0-function was introduced in ( * ? ? ?
* section 4 ) , and then , a more general theory of @xmath89-adic multiple @xmath0-functions has been developed in @xcite . to prove theorem [ thm4 ] ,
the following double zeta - function of hurwitz - lerch type was introduced : @xmath90 where @xmath4 , @xmath91 , @xmath48 with @xmath49 , @xmath50 .
this is a generalization of both and .
( frankly speaking , this is almost equivalent to consider with @xmath92 , @xmath93 . ) since @xmath73 can be written as a linear combination of functions of the form , it is sufficient to show a functional equation for
. this can be done in a way similar to the proofs of theorem [ thm1 ] and theorem [ thm3 ] ( see ( * ? ? ?
* sections 5 , 6 ) ) .
let @xmath94 be a sequence of complex numbers .
in @xcite , the authors considered the double series @xmath95 the original motivation of @xcite is to study the case when @xmath96 s are fourier coefficients of modular forms . in @xcite @xcite , manin extended the theory of periods of modular forms replacing integration along geodesics by iterated integrations . in this context some kind of multiple dirichlet series with fourier coefficients on the numerator naturally appears ( cf .
@xcite ) . multiple series with fourier coefficients is also expected to be useful to evaluate certain multiple sums of fourier coefficients ( analogously to de la bretche @xcite , ishikawa @xcite , essouabri @xcite etc ) . in the present paper
we consider a slight generalization of , that is @xmath97 where @xmath92 with @xmath93 .
this is a generalization of both and .
( the exponential factor @xmath98 on the numerator of is to be included in @xmath96 . )
the following two theorems ( theorems [ thm5 ] and [ thm6 ] ) are the main results of the present paper . when @xmath99 , those theorems are theorem 1.1 and theorem 1.5 in @xcite , respectively . concerning the order of the coefficients @xmath96 , we assume : \(i ) @xmath100 with a certain constant @xmath101 , where @xmath102 is an arbitrarily small positive number .
then , the dirichlet series @xmath103 is absolutely convergent in the region @xmath104 , and we further assume \(ii ) can be continued meromorphically to the whole complex plane with only finitely many poles . under these two assumptions , we see that is absolutely convergent in the region @xmath105 ( using ( i ) and ( * ? ? ?
* theorem 3 ) ) . as a generalization of ,
define @xmath106 where @xmath107 then we have [ thm5 ] assume ( i ) and ( ii )
. then the functions @xmath108 and @xmath109 can be continued meromorphically to the whole space @xmath28 , and the functional equation @xmath110 holds . in particular ,
applying theorem [ thm5 ] to the special case @xmath111 for only one fixed @xmath112 , and @xmath113 for all other @xmath112 , we obtain a functional equation for the single series @xmath114 which gives the `` refinement '' ( or the `` decomposition '' ) of theorem [ thm5 ] in the sense of ( * ? ? ?
* remark 1.4 ) .
note that is a generalization of the series @xmath115 ( @xmath116 ) which was used in @xcite and @xcite . also ram murty and sinha @xcite studied analytic properties of this type of function and its generalizations .
now we consider the case when @xmath96 s are fourier coefficients of modular forms .
let @xmath117 be the complex upper half plane , and @xmath118 where @xmath119 .
when @xmath120 is a holomorphic cusp form of even weight @xmath121 with respect to the hecke congruence subgroup @xmath122 , it is well known that assumptions ( i ) and ( ii ) are satisfied .
therefore , obviously , theorem [ thm5 ] can be applied to this case . in this case , however , we can show a different type of functional equation , using modular relations .
let @xmath123 and denote its fourier expansion by @xmath124 .
define @xmath125 write @xmath126 and @xmath127 instead of @xmath128 and @xmath108 , respectively .
then we have [ thm6 ] assume ( i ) and ( ii ) , and @xmath120 is as above .
then the functions @xmath129 can be continued meromorphically to the whole space @xmath28 , and the functional equation @xmath130 holds .
[ rem3 ] there is only one factor @xmath96 on the numerator of , .
this is an unsatisfactory point ; it is desirable to study the double series with two factors ( such as @xmath131 ) on the numerator .
however our present method can not be applied to such a situation . in the remaining sections we give the proofs of theorems [ thm5 ] and [ thm6 ] .
since the proofs are straightforward generalization of the proofs described in @xcite , we just outline the argument briefly .
denote @xmath132 and @xmath133 .
_ step 1_. the integral @xmath134 converges in the region @xmath135 , @xmath136 , @xmath137 , and @xmath138 this can be shown by using and the taylor expansion @xmath139 _ step 2_. let @xmath140 and divide the integral @xmath141 as @xmath142 say .
the important point here is that the function @xmath143 is holomorphic at @xmath144 .
the idea of this type of decomposition goes back to motohashi s short note @xcite , which influences @xcite ( see remark [ rem1 ] ) .
_ step 4_. using estimates of @xmath143 ( @xcite ) and @xmath147 ( ( * ? ? ?
* lemma 4.3.3 ) ) , we find that the integral @xmath148 is convergent in the region @xmath136 . since implies @xmath136 ,
now we know that the decomposition is valid in the region
. _ step 5_. changing the path of the inner integral of @xmath148 by the contour @xmath149 consisting of the half - line on the positive real axis from infinity to a small positive number , a small circle counterclockwise round the origin , and the other half - line on the positive real axis back to infinity .
then @xmath150 where @xmath151 since @xmath152 is convergent in the region @xmath153 and the right - hand side of can be continued to the whole space @xmath28 ( by assumption ( ii ) ) , now we see by that @xmath141 is ( and hence @xmath108 is ) continued meromorphically to the region .
_ step 6_. assume @xmath154 ( a little smaller than ) , and change the path @xmath149 on the right - hand side of by @xmath155 ( @xmath156 ) , and let @xmath157 .
counting the residues , we have @xmath158 where @xmath159 @xmath160 _ step 7_. applying to the right - hand side of and putting @xmath161 , we obtain @xmath162 where @xmath163 similarly we have @xmath164 the interchanges of summation and integration in steps 6 and 7 can be verified similarly to the argument in ( * ? ? ?
* section 4 ) , in the region . substituting and into ( and then ) , we obtain @xmath165 where @xmath166 _ step 9_. the final task is to show the meromorphic continuation of the functions @xmath108 and @xmath109 .
the continuation of @xmath30 can be done by using the asymptotic expansion of the confluent hypergeometric function . then gives the continuation of @xmath169 to the whole space @xmath28 .
this completes the proof of theorem [ thm5 ] .
let @xmath120 be as in the statement of theorem [ thm6 ] .
then @xmath147 is of exponential decay as @xmath170 , and hence the integral is convergent in the region @xmath171 ( wider than ) .
therefore , the formulas stated in steps 6 and 7 in the preceding section are now valid under the simple assumption @xmath172 . applying the modular relation to and , and
changing the order of integration and summation , we obtain @xmath173 @xmath174 similarly to ( * ? ? ?
* section 6 ) , we can verify the above argument in the region @xmath175 from , and , we obtain in the region .
therefore the only remaining task is to prove the meromorphic continuation of @xmath176 .
we only discuss the case of @xmath177 . substituting the expression of the confluent hypergeometric function into the right - hand side of , putting @xmath178 and changing the order of summation and integration , we obtain @xmath179 where @xmath180 satisfies the conditions @xmath181 and @xmath182 .
putting @xmath183 we obtain @xmath184 where @xmath185 . as a generalization of (
* lemma 7.1 ) , we can show @xmath186 for @xmath187 with @xmath188 .
therefore by the mellin inversion formula we have @xmath189 where @xmath190 and the path of integration is the vertical line @xmath191 . substituting into and changing the order of integration we obtain @xmath192 using the beta integral formula
we can show that the inner integral on the right - hand side of is @xmath193 and hence @xmath194 finally , modifying the path of integration on the right - hand side of in the same way as in ( * ? ? ?
* section 7 ) , we can show the meromorphic continuation of @xmath177 .
the case of @xmath195 is similar .
the proof of theorem [ thm6 ] is complete .
furusho , h. , komori , y. , matsumoto , k. and tsumura , h. , desingularization of complex multiple zeta - functions , fundamentals of @xmath89-adic multiple @xmath0-functions , and evaluation of their special values , preprint .
( arxiv:1309.3982 ) komori , y. , matsumoto , k. and tsumura , h. , functional equations and functional relations for the euler double zeta - function and its generalization of eisenstein type , _ publ .
debrecen _ ,
* 77 * ( 2010 ) , 15 - 31 .
matsumoto , k. , asymptotic series for double zeta , double gamma , and hecke @xmath0-functions , _ math .
cambridge phil .
_ , * 123 * ( 1998 ) , 385405 ; corrigendum and addendum , _ ibid .
_ , * 132 * ( 2002 ) , 377384 .
matsumoto , k. , on the analytic continuation of various multiple zeta - functions , _ number theory for the millennium ii _
, proc . millennial conf . on number theory ,
bennett , m. a. et al .
( eds . ) , a k peters , 2002 , pp.417 - 440 .
murty , r. and sinha , k. , multiple hurwitz zeta functions , _ multiple dirichlet series , automorphic forms , and analytic number theory _
, d. bump et al .
( eds . ) , _ proc .
pure math .
_ * 75 * , amer .
, 2006 , pp.135 - 156 . | we first survey the known results on functional equations for the double zeta - function of euler type and its various generalizations .
then we prove two new functional equations for double series of euler - hurwitz - barnes type with complex coefficients .
the first one is of general nature , while the second one is valid when the coefficients are fourier coefficients of a cusp form . |
it is a remarkable fact that the entropy of a black hole can be understood in terms of the surface area of its horizon @xcite .
why is there such a connection between the size of the hole and its entropy ? in this paper
we make a small observation that might be relevant to this connection .
let us consider extremal holes , using string theory .
consider first the 3-charge system , which can be made with charges d1-d5-p wrapped on the cycles of a compactification @xmath0 @xcite . at weak coupling @xmath1 we can count the microstates that have charges @xmath2 , getting an entropy @xmath3 .
next , follow the system towards larger coupling @xmath4 .
now we can say that these charges should describe an extremal black hole in 4 + 1 noncompact dimensions . a classical black hole of this type has an infinitely deep throat , ending in a horizon whose area gives @xmath5 ( fig.[first](a ) ) .
but because the throat is infinitely deep , the system has no energy gap ; we can place a wavepacket in the throat as deep as we want , which means that we can get as much of a redshift as we want .
this sounds strange , since we are still dealing with a finite sized system made of a finite number of charges , so we would have expected that the energy gap to be small but nonzero . in @xcite it was argued that as we increase @xmath4 from zero , the ` size ' of the brane bound state increases as well , such that at all values of @xmath4 the size of the state is of the same order as the horizon radius at that @xmath4 .
the essential reason for this large size was the phenomenon of ` fractionation ' , which makes the brane bound state have a size that depends on the number of charges it contains , instead of being a fixed size like planck length or string length .
the computation of @xcite was a very crude estimate .
but in @xcite it was noted that one can start with the 2-charge system , which gives the simplest model of a black hole . in this case
all states can be understood explicitly in the cft and in gravity , and the 2-charge bound states were found to be ` horizon sized fuzzballs ' .
next , we find that we can take very special states of the d1-d5-p system where many quanta are chosen to lie in the same mode .
this makes the state ` classical ' , and we can construct the geometry for these states .
one finds the situation in fig.[first](b ) , where the throat ends in a smooth cap rather than a horizon @xcite .
the energy gap is now finite , and is found to agree exactly with the gap in the dual cft state .
the generic state will not have all its quanta placed in a few modes , so we will not get a classical geometry ; rather the throat will end in a very quantum ` fuzzball ' ( fig.[first](c ) ) . instead of the precise energy ` gap ' mentioned above we can consider a closely related quantity : the depth of the throat up to the quantum region .
an indirect argument shows that this depth again agrees with the corresponding quantity in the cft @xcite .
large families of 3 charge geometries ( in 4 + 1 dimensions ) and 4 charge geometries ( in 3 + 1 dimensions ) have now been constructed @xcite .
these all have the same charges and mass as the extremal black hole , but no horizon ; the throat is ` capped ' in each case .
if the picture fig.[first](c ) is a correct description of the hole , then black holes are not that different from other quantum systems ; they are large quantum balls with information distributed throughout a horizon sized region .
what basic physics governs the size of these objects ?
a principal feature of black holes is their large entropy .
but entropy is just the log of the volume of phase space occupied by the allowed states of the system .
if this phase space is large , perhaps it is not possible to make the system too small in size , and this may be the reason that the generic state becomes a horizon sized ` fuzzball ' .
we will explore this notion with the help of a small calculation .
consider the 2-charge system , where we compactify iib string theory on @xmath6 .
we wrap an ns1 string @xmath7 times along @xmath8 , and add @xmath9 units of momentum along @xmath8 .
the entropy is large since it grows with the charges s_micro=2 [ en1 ] now let us picture the states contributing to this entropy . the ns1 is a ` multiwound string ' wound @xmath7 times around the @xmath8 before joining back to itself . in fig.[intro](a ) , at the top , we have drawn this string opened up to its full length , ( i.e. , we have taken the @xmath7-fold cover of the @xmath8 ) .
the momentum is carried by transverse vibrations of the ns1 , and fig.[intro](a ) depicts a ` generic state ' contributing to the entropy ( [ en1 ] ) , with its corresponding transverse amplitude . below this string
we sketch the trajectory taken by the ns1 in the noncompact directions ; we see that the transverse vibrations of the string cause it to spread over a ball in the these noncompact directions . below this ball
we sketch the geometry produced by the string state .
it is flat space at infinity , then a throat narrowing towards the location of the string , but this throat does not reach down to a pointlike singularity .
instead , the throat ` caps off ' when we reach down to the boundary of the ball .
we have taken a classical profile for the ns1 , but if we include quantum fluctuations , then the region inside this boundary gives the ` fuzzball ' region .
it was observed in @xcite that the area of the boundary of this ball satisfies ~~s_micro [ rel ] thus the actual microstates of the system have no horizon , but the boundary of the region where microstates differ from each other satisfies a bekenstein type relation with the entropy inside the boundary .
can this be an instance of a more general principle ?
in @xcite the rotating 2-charge system was also considered ; here the microscopic entropy is @xmath10 , and boundary area of the fuzzy region around the ring again satisfied the analogue of ( [ rel ] ) . to explore further this relation between boundary area and enclosed entropy , we proceed as follows .
we again consider the ns1-p system , but restrict our states to a ` sub - ensemble ' , which we now describe .
the total momentum carried on the string is fixed by the momentum charge that we have chosen .
but we can have states where the momentum is in high wavenumbers with small transverse amplitude , or in low wavenumbers with large transverse amplitude .
the entropy is optimised for some typical wavenumber and its corresponding amplitude , and such a typical state has been depicted in fig.[intro](a ) . but
now let us restrict to the subset of states which have a wavenumber higher than this optimal value ; the amplitude of transverse vibration will then be smaller than the amplitude in the generic state ( fig.[intro](b ) ) .
the microscopic entropy of states restricted to this smaller amplitude is smaller than ( [ en1 ] ) by a factor which we call @xmath11 . in the spacetime description
, the ns1 now lives inside a smaller ` ball ' , and the spacetime geometry has a throat which becomes narrower and deeper before capping off .
we compute the boundary of the fuzzball region in this case , and ask if , to leading order , it has also decreased by a factor @xmath11 . we find that such is indeed the case .
we repeat the computation for the case with rotation , and find a similar agreement .
this is the main computation of this paper .
the sub - ensembles with smaller entropy had their fuzzball boundaries at deeper points in the throat , where the transverse area was smaller .
we observe that we can recast this computation in a form where we hold the location of the boundary fixed but reduce the _ distortion _
@xmath12 allowed in the metric at the boundary . allowing a smaller distortion forces the throat to continue to be the ` naive ' one to deeper depths , and thus leads to the sub - ensemble with small entropy .
we suggest a relation between the ` capped ' nature of throats and an old puzzle with extremal holes : in a euclidean gravity computation one finds that the entropy of an extremal hole is _ zero _ @xcite .
the euclidean computation uses an infinitely long uniform throat , which does not correspond to a state in the fuzzball picture .
string theory counts actual microstates , and these all ` cap off ' before reaching the end of the infinite throat .
finally , we check that quantum corrections will not change our conclusions . to this end
we list the sources of quantum effects , and examine their roles . in the process
we compute various length and curvature scales to obtain a picture of the generic 2-charge extremal state for our @xmath6 compactification .
we end with some conjectures on how phase space effects may help resolve puzzles associated with black holes .
let us recall the basic structure of the 2-charge extremal system .
the results summarized below can be found in more detail in the review @xcite . take type iib string theory , with the compactification m_9,1m_4,1_4s^1
the 4-manifold @xmath13 can be @xmath14 or k3 ; we will let it be @xmath14 for now .
let this @xmath14 have volume @xmath15 , and let the @xmath8 have length @xmath16 .
the string coupling is @xmath17 , and we will set @xmath18 . now consider the bound state made with charges ns1-p , where the ns1 charge corresponds to an elementary string wound @xmath7 times on @xmath8 , and the p charge is @xmath9 units of momentum along @xmath8 .
the ` naive ' supergravity solution for these charges is ds^2_string&=&h[-dudv+kdv^2]+_i=1 ^ 4 dx_idx_i+_a=1 ^ 4 dz_adz_a + b_uv&=&-12[h-1 ] + e^2&=&h + h^-1&=&1+q_1r^2 , k = q_pr^2 [ naive ] where @xmath19 , the @xmath20 are coordinates along the @xmath14 and @xmath21 are coordinates in noncompact space .
we have q_1=g^2v n_1 , q_p = g^2v r^2 n_p [ q1qp ] as shown in @xcite , the actual geometries produced by the states of this system are different from the ` naive ' solution ( [ naive ] ) .
the p charge on the ns1 is carried by transverse vibrations of the ns1 .
we will focus on the vibrations in the noncompact directions @xmath21 as these will give us an estimate for transverse size of the bound state .
were studied in @xcite and the fermionic excitations were studied in @xcite . ]
these vibrations are described by a ` profile function ' @xmath22 .
the ns1 string , in the bound state , is a single ` multiwound ' string of total length l_t=2r n_1 [ tlength ] the vibration profile is a function of @xmath23 . there are four components of this displacement , @xmath24 , collected into the vector @xmath25 . thus the noncompact vibrations are described by @xmath22 with @xmath26 ranging over @xmath27 .
the geometry for a profile @xmath22 is ds^2_string&=&h[-dudv+kdv^2 + 2a_i dx_i dv]+_i=1 ^ 4 dx_idx_i+_a=1 ^ 4 dz_adz_ab_uv&=&-12[h-1 ] , b_vi = ha_ie^2&=&h [ ttsix ] where h^-1&=&1+q_1l_t_0^l_t + k&=&q_1l_t_0^l_t + a_i&=&-q_1l_t_0^l_t [ functionsq ] note that for @xmath28 we have k[_0^l_t ] so that we identify q_p = q_1l_t_0^l_t(f(v))^2[q1p ] we can perform a sequence of s and t dualities , and map this ns1-p system to the d1-d5 system .
now we have d1 branes wrapped on @xmath8 , and d5 branes wrapped on @xmath6 .
we will use ` primed ' coordinates when we are in the d1-d5 duality frame .
the d1 , d5 charges are @xmath29 with @xmath30 the metric and dilaton are ds^2_string=[-(dt-a_i dx^i)^2+(dy+b_i dx^i)^2]+dx_idx_i+dz_adz_a [ qsix ] e^-2=h_5h_1 where the harmonic functions are h_5&=&1+^2q_1l_t_0^ l_t dv|x-f(v)|^2h_1&=&1+^2q_1 l_t_0^ l_t dv ( f(v))^2|x-f(v)|^2 , + a_i&=&-^2q_1l_t_0^ l_t dv f_i(v)|x-f(v)|^2 [ functionsqq ] here @xmath31 is given by db=-*_4da [ vone ] and @xmath32 is the duality operation in the 4-d transverse space @xmath33 using the flat metric @xmath34 . in the duality from ( [ ttsix ] ) to ( [ qsix ] )
we have kept the metric to be unity at infinity in the string frame .
this has led to a scaling of the coordinates @xmath35 x = x [ xpx ] with and the moduli in the ns1-p frame is given in @xcite . ]
= g r [ mu ] by contrast the ` naive ' geometry which one would write for d1-d5 is ds^2_naive=1[-dt^2+dy^2]+dx_idx_i+dz_adz_a [ naived ] where @xmath36 .
writing @xmath37 we see that ( [ naived ] ) is the ` naive ' metric that is conventionally written for the d1-d5 system : it is flat space at infinity , it has a ` throat ' for @xmath38 , ending in a singularity at @xmath39 .
the geometry of the throat is the poincare patch of @xmath40 ( with a periodic identification @xmath41 ) . in
the ` actual ' geometries ( [ qsix ] ) the throat is ` capped off ' smoothly at small @xmath42 , and the information of the configuration is in the shape of the cap .
for later use , we note that q_1=^2 q_p = gvn_1 , q_5= ^2 q_1=g n_5 [ q1q5 ]
while the geometries ( [ qsix ] ) have been generated by starting with a classical profile @xmath22 for the ns1 string , the actual state of the system will correspond to a quantum state of the string and will be a ` fuzzball ' .
we will discuss the various sources of quantum effects in section [ nature ] below , but for now we are just interested in the location of the ` boundary ' of this fuzzball ; i.e. the surface inside which the typical state begins to differ from the naive geometry ( [ naived ] ) . in @xcite
it was observed that the area of this boundary obeyed ~~s_micro in this section we review this result ; more details can be found in @xcite . in the next section
we will find that a similar result holds for the sub - ensembles that we consider .
consider the ns1-p system .
the large entropy of the system arises from the fact that the momentum p on the ns1 can be partitioned in many ways among different harmonics . since we are dealing with bps states
, we can look at them at weak coupling and then follow them to strong coupling . at weak coupling
we have a string in flat space carrying transverse vibrations .
the total energy on the string is e = p = n_pr=2n_1n_pl_t [ tenergy ] each excitation of the fourier mode @xmath43 carries energy and momentum e_k = p_k=2kl_t let there be @xmath44 excitations of type @xmath45 in the harmonic @xmath43 ( the index @xmath45 runs over 8 bosonic excitations and 8 fermionic excitations . )
then we must have _ i _ k k m^i_k = n_1n_p [ equationm ] we assign each bps state the same probability , since it has the same charges and mass ( later we will add an additional weighting with angular momentum ) .
the probability distribution is then a thermal one , describing left moving massless quanta in a box of length @xmath46 , with temperature t=^-1 = [ 12 el_t c]^12=l_t [ temperature ] ( here the central charge @xmath47 describes the effective number of degrees of freedom of the massless gas ; the bosons contribute @xmath48 and the fermions @xmath49 . )
the entropy is s_micro=2 [ entropy ] consider the bosonic excitations . since we will only be estimating quantities upto factors of order unity , we restrict to just one flavor of boson and ignore the index @xmath45 . for this boson
the occupation number of the @xmath43th harmonic is = 11-e^-e_k=11-e^-k [ occupation ] let us build up a rough picture of how energy is distributed among harmonics in the generic state . since we just have a thermal gas of excitations , the typical excitation quantum will have energy @xmath50 , which from ( [ temperature ] ) gives for the mean harmonic order |k~ [ bark ] from ( [ occupation ] ) we see that for modes with @xmath51 the occupation number is ~1 [ unity ] the number of harmonics that have @xmath51 is itself of order @xmath52 . for concreteness let us consider a range |k < k < 32|k [ range ] which has a width k~|k ~ the number of quanta in this range is n~k m_k ~ [ nvalue ] with each quantum carrying energy @xmath53
the total energy is e~_|k n~2l_t|k n~2l_t n_1n_p which agrees with ( [ tenergy ] ) . consider fig.[intro](a ) . in the next subsection we will find the radius of the ball enclosing the curve traced out by the ns1 .
but there are many subtleties in defining this ball .
we will mention some of these here , and then specify how we will define the transverse size of the state . in the top sketch of fig.[intro](a )
we have shown the ns1 carrying transverse oscillations of some typical wavelength @xmath54 .
if we had a fourier wave of one precise wavelength @xmath54 , then the transverse size would be @xmath55 .
if we have a spread over fourier modes , and the typical wavenumber is @xmath56 , then in a typical oscillation the string will have a transverse displacement @xmath57 . the ball in fig.[intro](a ) is drawn with the assumption that the oscillations of the string keep the string going back and forth in a region with radius of this order , and we will use this picture to compute the ball size below .
thus we are in a sense taking the ` smallest possible size ' allowed to the ball when the vibrations have mean wavenumber @xmath56 .
but there are many other measures possible for the ` size ' of the state .
for example , when there is a spread over @xmath43 we can compute many different ` averages ' , which weight low and high @xmath43 in different ways . in @xcite a different way of computing
` size ' was used and a much larger radius for the ball was found .
this does not conflict with the essential idea of fuzzballs , since if the ball size is bigger than horizon size the information can not be trapped inside a horizon .
but we would like to see which way of measuring size gives a relation to the area of the boundary , and as found in @xcite the definition we will use gives a direct relation to the idea of area entropy . to see how other definitions of size might be related to the one we will choose ,
consider again the string in fig.[intro](a ) . during each wavelength of oscillation
the transverse displacement is @xmath58 .
suppose we say that the path of the string in the noncompact directions is a random walk , with step size @xmath58 .
then we will get a much larger ` ball size ' than @xmath58 , since the random walk has @xmath59 steps and will drift over a large transverse distance .
but if we fourier transform this path , we will see that this large drift corresponds to low harmonics on the string , i.e. , modes with @xmath60 .
in @xcite it was noted that the ` tail ' of low frequency modes in the distribution makes the entropy from the boundary area larger than the microscopic entropy by a logarithmic factor .
we will filter out the effects on size that arise from the tail of low harmonics .
but because of these uncertainities in defining the ensemble , _ we will ignore all log corrections below .
_ recall that the total length of the ns1 string is @xmath46 given by ( [ tlength ] ) .
thus for a mode in harmonic @xmath56 the wavelength of vibrations is = l_t|k~2r n_1~r [ wavelength ] we wish to ask how much the transverse coordinate @xmath35 changes in the process of oscillation .
thus we set @xmath61 , and find |x|~||y~||r [ step1 ] from ( [ q1p ] ) we know that q_p ~q_1 ||^2 which gives | x|~r~1 [ ffone ] where we have used ( [ q1qp ] ) .
we wish to compute the ` size ' of a generic fuzzball state .
it turns out that it is easier to understand quantum corrections in the d1-d5 frame than in the ns1-p frame @xcite .
thus we will work in the d1-d5 frame below . in this duality frame the ` spread ' ( [ ffone ] ) of the state in coordinate terms is r|x|=|x|~=g r [ bvalue ] where we have used ( [ mu ] ) .
for |x||x| we have and the solution becomes approximately the naive geometry ( [ naived ] ) .
we see that the metric settles down to the naive metric outside a certain ball shaped region @xmath62 .
let us now ask the question : what is the surface area of this ball ? to compute this , we will take the naive metric ( [ naived ] ) , put a surface at @xmath63 , and find the area of this surface . since we will need a more general result in a moment ,
let us set r=b and compute the area of the bounding surface in the naive geometry ( [ naived ] ) . in the bekenstein relation we can compute the area of the horizon surface in the 10-d einstein metric and divide by @xmath64 , or compute the area in the 5-d einstein metric and divide by @xmath65 = a_e^(5)g_n^(5 ) [ newton ] we will compute the area in 10-d , where we have an 8-dimensional surface : an @xmath8 from the direction @xmath66 , a @xmath14 from the directions @xmath67 , and an @xmath68 from the angular directions in @xmath69 .
we first compute in the string frame , using the metric ( [ qsix ] ) , and then convert to the einstein frame . in the string frame
the @xmath8 has a length l_y=(2r ) ( h_1h_5)^-14 [ length1 ] the angular sphere in the @xmath69 directions gives v_s^3=2 ^ 2 b^3 ( h_1h_5)^34 the @xmath14 gives v_4=(2)^4 v ( h_1h_5 ) converting to the einstein metric @xmath70 from the string metric @xmath71 is done through g_e= e^-2g_s=(h_1h_5)^-14g_s [ convert ] with all this we find the area of our surface in the einstein metric a_e^(10)=12 ( 2)^7 ( h_1h_5)^12 b^3 v r writing h_1 , h_5 we get a_e^(10)=12 ( 2)^7 ( q_1q_5)^12 v r b finally , using g_n^(10)=8 ^ 6 g^2 we find ~(q_1q_5)^12 v rg^2
b [ areafinal ] where we have now dropped all numerical constants since we can specify the value @xmath72 only upto a factor of order unity . setting @xmath73 to the value ( [ bvalue ] ) we get ~(q_1q_5)^12 v rg^2 ( g r)~ where we have used ( [ q1q5 ] ) . noting ( [ entropy ] )
, we now observe that @xcite ~s_micro [ relation ] thus we see that the size of the ` fuzzball ' made by the generic state is such that its surface area satisfies a bekenstein type relation with the entropy of the fuzzball . ) was derived in @xcite using arguments based on a ` stretched horizon ' . ]
it was argued in @xcite that if a quantum falls into the region enclosed by this boundary , the complicated nature of the interior state traps the quantum for very long times , so that this boundary behaves like a ` horizon ' for observers that are observing the system only over classical timescales : i.e. times of order the radius of the @xmath68 .
in this section we will perform the main computation of this paper .
we wish to get some understanding of why the relation ( [ relation ] ) turned out to be true .
as discussed in the introduction , black holes have a large entropy , and entropy is a measure of phase space volume .
one might therefore conjecture that the ` fuzzball ' can not be too small because it needs to contain enough phase space to account for this large entropy .
note that the area @xmath74 in ( [ relation ] ) is ` large ' in terms of the elementary scales in the theory . for the d1-d5 system ,
the dilaton approaches a constant down the ` throat ' of the geometry , so string length @xmath75 and the 10-d planck length @xmath76 are of the same order everywhere . in other words , the ratio @xmath77 does not depend on the charges , which we will take to satisfy @xmath78 . since ( [ relation ] ) says that the horizon area in planck units gives the entropy @xmath79 , we see the 10-d horizon area is much larger than planck or string scale . from ( [ newton ] ) we see that the horizon area is large in terms of the 5-d planck scale as well .
how do we explore the role of phase space in determining the size of the fuzzball ? as explained in the introduction , we will take a subset of states which we call a ` sub - ensemble ' which contains states with transverse spread smaller than the spread of the generic state .
we can compute the area of the bounding surface for the typical state of this sub - ensemble ; this will be smaller than the area @xmath74 in ( [ relation ] ) by some factor @xmath80 ( with @xmath81 ) .
on the other hand , the entropy of such states will also be smaller than @xmath3 by some factor , and we wish to check if this factor is also @xmath80 .
we will find that the factors indeed agree ( up to log terms that we will ignore ) . in the next section
we will repeat the analysis for the 2-charge extremal ` black ring ' , which is the same 2-charge system but with rotation .
we will again find a similar agreement .
let us discuss the microscopic description first .
as mentioned above , we will be working with the d1-d5 duality frame later , so we will make the replacement n_1 n_5 , n_p n_1 in all relations which had been developed in the ns1-p duality frame .
consider the entropy ( [ entropy ] ) .
we want to get a crude argument for why this entropy is order @xmath82 in the charges . from ( [ bark ] ) we see that the generic quantum is in harmonic @xmath83 , and from ( [ nvalue ] ) we note that there are @xmath84 such quanta . from ( [ unity ] )
we have that @xmath85 for these modes @xmath43 .
let us thus take the crude approximation that this occupation number can be @xmath86 or @xmath87 for each mode in the range ( [ range ] ) , with equal probability .
the number of possibilities is of order ~2^ so that the entropy is s_micro ~ we could certainly make a much better estimate if we wished ; for example we can take into account the fact that each mode has several polarization states , or the fact that the occupation numbers are not just @xmath88 , or the fact that occupation probability is not uniform in the range ( [ range ] ) . but the above crude estimate is enough for our purposes , since we will extend it to the case that we now wish to cover .
we are interested in looking at the subset of states where the typical harmonic order is not @xmath89 but rather |k_m~m [ knew ] for some @xmath90 . since the total energy is still ( [ tenergy ] ) the number of excitation with @xmath43 of this order will be n_m~m [ excitations ] as before , we interpret ( [ knew ] ) as saying that @xmath43 lies in a range m < k <
32 m [ rangenew ] so that the range of @xmath43 has a width k_m ~m note however that this time the number of excitations @xmath91 in ( [ excitations ] ) is much less than the number of elements in the range ( [ rangenew ] ) ~1m^2 < < 1 [ less ] so our task is to count the number of ways that @xmath92 excitations can be distributed in the range @xmath93 .
because of ( [ less ] ) , we see that the @xmath92 excitations will be sparsely distributed in the range @xmath93 , so we do not have to worry about more than one excitation falling on the same value of @xmath43
. each excitation can be in one of @xmath93 possible values of @xmath43 , giving @xmath94 , and we must then divide by @xmath95 since we can permute the @xmath92 excitations without changing the state .
we thus get for the number of possibilities ~1n_m!(k_m)^n_m thus the entropy is s_micro , m&= & n_m - [ n_mn_m - n_m]&= & n_m
[ + 1]&~ & n_m [ + 1 ] & ~ & m [ + 1 ] where in the third line we have used ( [ less ] ) and in the last line we have used ( [ excitations ] ) . as mentioned above
, we will drop all log corrections ; we will just note that the power of @xmath11 by which @xmath96 changes is the power @xmath97 .
so we just write s_micro , m~m [ mfinal ] let us now turn to the gravity description of the states in our sub - ensemble . instead of the wavelength @xmath54 in ( [ wavelength ] )
we have the wavelength _
m = l_t|k_m~1 m r [ wavelengthp ] following the same steps as in eq.([step1])-eq.([ffone ] ) we get ( in place of ( [ ffone ] ) ) |x|~1 m [ ffonep ] thus the corresponding coordinate size in the d1-d5 frame is ( using ( [ xpx ] ) ) |x|=1mg r [ bvaluenew ] we must therefore set b=1mg r in ( [ areafinal ] ) . calling the area of the boundary @xmath98 we get ~(q_1q_5)^12
v rg^2 ( 1mg r)~1 m comparing with ( [ mfinal ] ) we see that ~s_micro , m this is the main computation of this paper .
we will next see that a similar result of obtained for the system with rotation .
we will comment on physics of these results in section [ what ] below .
we can give our 2-charge extremal system an angular momentum , and obtain a 2-charge _ black ring_. in @xcite this system was also considered , and it was again found that the area @xmath99 of the boundary of the fuzzy region gave @xmath100 .
we would like to repeat for this system the steps of the previous section ; i.e. , we restrict to a sub - ensemble with smaller entropy , and then look for the area of the bounding surface for this sub - ensemble .
we will again find an agreement similar to the one for the non - rotating system .
the ns1 describes a path close to a circle ; the tube shaped boundary encloses the region inside which the typical curve fluctuates ( b ) the paths for the sub - ensemble .
the vibrations have smaller amplitude and higher wavenumber , and the tube is therefore ` thinner ' with a smaller surface area.,width=480 ] we begin by recalling the computation of @xcite .
let us start with the ns1-p frame .
the charges are still @xmath101 , but we require the system to have an angular momentum @xmath102 . in @xcite it was shown that @xmath103 , so we will set j = n_1n_p = n_1n_5 , 0<1 to give the system an angular momentum in the @xmath104 plane , we let the ns1 string have a profile ( v)=a e_1+a e_2+x(v ) [ residual ] the path traced out by the ns1 now lies close to a circle in the @xmath104 plane ; this circle has radius @xmath105 .
since we will be working with the d1-d5 system , we note the radius of this circle in the coordinates @xmath69 a== consider the ns1-p state .
we again have to put momentum excitations on the ns1 satisfying ( [ tenergy ] ) , but with the constraint that the state carries an angular momentum @xmath102 .
the angular momentum can be carried by the bosonic excitations as well as the fermionic excitations .
but if @xmath102 is large enough to be of order @xmath106 , we find that there is a simplification
. a large number of bosonic excitations drop to the lowest harmonic @xmath107 and carry all the angular momentum @xmath102 ; these modes therefore contribute nothing to the entropy .
the rest of excitations distribute themselves among all harmonics @xmath43 to yield the entropy , but they are not constrained by the requirement of carrying any net angular momentum .
the @xmath102 excitations carrying the angular momentum have energy e^rotation=2l_tj [ erot ] so the remaining modes carry energy e^vibrations=2l_t ( n_1n_p - j ) [ evib ] the entropy is therefore @xcite s_micro=2=2 [ ringm ] let us also note the nature of the generic state .
the energy ( [ evib ] ) has to be distributed between different harmonics .
thus we will have , in place of ( [ bark ] ) , a mean harmonic order |k~ [ kj ] the number of quanta with wavenumber of this order will be ( in place of ( [ nvalue ] ) ) n~ consider the path of the ns1 in the @xmath35 space .
we depict this in fig.[ring](a ) .
the central circle is the circle of radius @xmath108 , and the residual vibrations @xmath109 in ( [ residual ] ) make the actual ns1 profile wiggle around in a thin tube shape region around this circle .
we draw a tube shaped boundary to bound the region occupied by the ns1 in a typical state .
the area of this boundary ( in the @xmath69 plane ) will therefore be given by the area of an @xmath110 transverse to the circle , times the length of an @xmath8 along the circle direction .
we must first find the coordinate radius of this @xmath110 which measures the typical displacement of the ns1 from its mean location on the circle .
this takes a little more work than in the non - rotating case . with @xmath56 given by ( [ kj ] )
the analog of ( [ wavelength ] ) is = l_t|k~ r n_1 [ wavelengthj ] let us go to a particular point on the ring . in the vicinity of this point , the vector @xmath22 can be decomposed into two parts f(v)=f_rot(v)+f_vib(v ) here @xmath111 is the part that comes from the first two terms on the rhs of ( [ residual ] ) ; thus it points along the ring and gives it its angular momentum @xmath102 .
the part @xmath112 comes from @xmath109 in ( [ residual ] ) .
we have |(v)|^2=|_rot(v)|^2+|_vib(v)|^2 + 2 _ rot(v)_vib(v ) we will assume that the randomness of the fluctuations @xmath109 gives _
rot(v)_vib(v)=0 we have q_p = q_1[(2l_t)^2 j+|_vib(v)|^2 ] thus |_vib(v)|==1n_1 r where we have used ( [ q1qp ] ) .
the transverse displacement of the profile from the central circle of the ring is therefore given by a computation analogous to ( [ step1 ] ) |x|~|_vib(v)|~1 where we have used ( [ wavelengthj ] ) .
thus the radius @xmath113 of the sphere transverse to the ring is of the same order as the transverse displacements ( [ ffone ] ) in the non - rotating case . transforming to the d1-d5 frame
, we have |x|=r_s^2 ~g r [ rfinal ] as in the non - rotating case , we switch to the d1-d5 frame for our computations
. we will again compute the area in 10-dimensions , and divide by @xmath64 . outside the tube shaped boundary
, the geometry will be close to a ` naive ' metric .
we construct this naive metric by taking the general metric ( [ qsix ] ) , and replacing @xmath114 by the distance of @xmath69 from the central circle @xmath115 .
we find that the functions in ( [ qsix ] ) are given by h_1&=&1 + , h_5=1+,a_i dx^i&= & ( x_2dx_1-x_1dx_2 ) , + f_0&=&^1/2 [ bigeq ] for @xmath69 far outside the tube shaped region in fig.[ring](a ) , the actual metric will agree closely with this naive metric .
the length of the @xmath8 in the direction @xmath66 is ( in the string metric ) l_y=(2r ) ( h_1h_5)^-14 the @xmath14 gives v_4=(2)^4 v ( h_1h_5 ) in the noncompact space @xmath69 , the boundary of the tube has an @xmath110 transverse to the ring and an @xmath8 along the ring .
the @xmath110 has area a_s^2=4(r_s^2)^2(h_1h_5)^12 let @xmath116 be a coordinate along the ring , normalized so that @xmath117 .
close to the ring , the field @xmath118 has only a component along @xmath116 . from ( [ bigeq ] )
we find a_z=- the metric along the ring is then g_zz=-1 a_z^2=1q_1q_5(2ar_s^2)^2 n_1n_5 ( n_1n_5-j ) thus the circle @xmath8 along the ring has length l_1=2a = ( h_1h_5)^-14r_s^2 taking the product of the contributions from @xmath119 , and converting to the einstein metric using ( [ convert ] ) , we arrive at a_e^(10)=(2)^7 v r r_s^2 [ areaj ] using ( [ rfinal ] ) , we get @xcite ~~s_micro where at the last step we have compared with ( [ ringm ] ) .
we will see that all computations for this case are very similar to those in the non - rotating case .
we have to hold fixed the conserved quantities of the ring the charges @xmath29 and the angular momentum @xmath102 and then look for a subset of states that define our sub - ensemble .
thus we will look at states where we still have the energy ( [ erot ] ) from the lowest harmonic carrying the angular momentum @xmath102 , but the energy ( [ evib ] ) from the remaining excitations are taken to be in harmonics of order |k_m ~m [ bkj ] with @xmath120 .
since the energy @xmath121 is unchanged , the number of excitations with @xmath43 of this order will be n_m~m [ excitationsj ] following the steps in section ( [ mic ] ) we find s_micro~m [ mfinalj ]
fig.[ring](b ) shows the nature of our sub - ensemble . by letting @xmath121
be carried by harmonics @xmath122 instead of @xmath123 we have put the energy in vibrations that have smaller amplitude and higher wavenumber . because of the smaller amplitude the vibrations stay closer to the central circle of the ring .
thus the @xmath110 surrounding the ring will have a smaller area , while the length of the direction along the ring will remain unchanged .
putting @xmath124 ( ( eq.[bkj ] ) ) in ( [ wavelengthj ] ) instead of @xmath56 , and following the steps to ( [ rfinal ] ) we find r_s^2 ~1mg
r [ rfinalj ] putting this value of @xmath113 in ( [ areaj ] ) we get ~1m~s_micro where at the last step we have compared with ( [ mfinalj ] ) .
black holes have an entropy given by the area @xmath99 of their horizon .
but this fact does not tell us _
why _ the area is a measure of the entropy .
we can not answer this question fully at this stage .
but in the above sections we have tried to probe the relation between entropy and area more deeply , by looking at sub - ensembles of the full space of states .
let us first summarize the physics that we have seen .
consider the 2-charge system in fig.[twocharge ] . in [ twocharge](a ) , the dashed boundary is placed at the location of the generic ` cap ' .
the size of the @xmath8 at the location of this boundary is estimated by setting @xmath125 ( eq . ( [ bvalue ] ) ) in the naive geometry ( [ naived ] ) .
since this boundary captures the generic state , the entropy of allowed states inside this boundary is @xmath126 .
fig.[twocharge](b ) depicts the situation with the sub - ensembles that we considered .
the naive geometry goes upto a deeper point , where the @xmath8 is smaller , and the boundary of the region where the caps differ is therefore smaller than the area in fig.[twocharge](a ) by some factor @xmath11 .
the number of ` caps ' that come after this depth of throat is smaller too , with an entropy @xmath127 . in fig.[twocharge](c )
we have let the geometry be the naive one upto a distance longer than the longest throat allowed for any microstate .
the @xmath8 boundary is now even smaller , and with this value of the boundary parameters there are no ` capped ' states possible . what physics is suggested by these observations ?
the fuzzball conjecture says that black holes are just large quantum balls , not fundamentally different from any other quantum system .
thus different microstates of the fuzzball must be orthogonal wavefunctions of the system .
one can imagine that it is not easy to fit too many orthogonal wavefunctions in a given region ( with a given energy ) .
the number of orthogonal states reflects the phase space volume occupied by the ensemble .
it is plausible that a large phase space volume implies , for the stationary states of our system , a large size in ordinary space @xmath35 . in that case
the large size of fuzzballs would be the result of the large phase space of states they describe .
but if this reasoning were correct , then we might expect a more general relation , which would say that _ any _ sub - ensemble of states with entropy @xmath3 would have to have a boundary area that satisfied @xmath128 .
this is what we found for our sub - ensembles ; in fact they happened to be ` optimally packed ' sets of states , since they achieved the maximal entropy allowed for their boundary area .
it would be interesting to check how these ideas work for 3-charge and 4-charge systems . at present
it is not clear to us how sub - ensembles should be made for this case .
suppose we take the sub - ensemble in fig.[twocharge](b ) but place the boundary at the same location as the boundary in fig.[twocharge](a ) ( shown by the upper ring in [ twocharge](b ) ) .
what will we see at this boundary ?
for the states in the full ensemble in fig.[twocharge](a ) , the geometry at the boundary @xmath129 is distorted by order unity compared to the naive geometry , since the caps start to differ significantly from each other at this point .
for the states in the sub - ensemble of fig.[twocharge](b ) , there is very little distortion at this location ; the distortions will come much further down the throat at @xmath130 .
if we place the boundary at @xmath129 in the geometry of fig.[twocharge](c ) then we will see virtually _ no _ distortion .
thus we see that there are two geometric quantities that we can consider at a boundary .
one is the area @xmath99 of this boundary .
the other is the width @xmath12 of distortions that we allow at this boundary .
we can now describe the ensembles in fig.[twocharge](a),(b),(c ) in two equivalent ways . in the first
, we place the boundary at the location where the distortions become order unity , and then use the area @xmath99 at this boundary to differentiate between the ensembles . in the second we keep the boundary at any fixed location , say @xmath129 , and change the allowed width @xmath12 . for the case in fig.[twocharge ] ( a ) we will have a width of order unity , in [ twocharge](b ) the width would be smaller , and in [ twocharge](c ) it would be virtually zero . as we will see in the next subsection , the second way of defining sub - ensembles may be related to other approaches to entropy like the use of asymptotic symmetries .
the principal feature that we see in fuzzballs is that each microstate has a different structure after some distance down the throat .
thus we see the ` hair ' responsible for the entropy of the system . with this explicit picture of the ` hair ' we might be able to understand better some other approaches to entropy : \(i ) the distortions at the fuzzball boundary give an _ entanglement _ @xcite between the states inside the cap and the geometry just outside . in the fuzzball picture ,
the latter geometry is not the featureless naive geometry outside a horizon , but reflects the distortions of the ` cap ' .
if we allow virtually no distortions , as in fig.[first](a ) , we get no states .
\(ii ) there is evidence of _ holography _
@xcite , since the entropy in the possible ` caps ' is controlled by the area @xmath99 and width @xmath12 of fluctuations at the _ boundary_. in a general statistical system there would be no reason for the boundary data to reflect the entropy in the interior .
\(iii ) it may be possible to relate the @xmath12 caused by the fuzzball caps to the algebra of diffeomorphisms studied in @xcite .
the central charge and level of this algebra appear to give a count of states , but it has never been explicitly clear what is being counted .
black hole entropy has also been derived as a noether charge for diffeomorphisms ( in the time direction ) @xcite , but again it is not clear what this charge is counting .
the fuzzball caps naturally generate distortions , and holography suggests that all information of the cap is captured by the distortions at the boundary .
thus any method of counting these distortions at the boundary should relate the entropy to the boundary properties .
the entropy of schwarzscild holes is given by the area of their horizon .
this is also true for holes with charge , but what about _ extremal _ holes , which have the maximum charge for their mass ?
the extremal reissner - nordstrom hole also has a nonvanising horizon area @xmath99 , but it has been argued that its entropy should be _ zero _ @xcite . in this argument one
notes that the extremal hole has an infinitely long throat and therefore a different topology from nonextremal holes . for nonextremal holes one must compactify the euclidean time direction @xmath131 with a given period @xmath132 to avoid a singularity at the horizon . for the extreme hole
we find that we can compactify @xmath131 with _ any _ period ; since the horizon is infinitely far away we do not get a singularity from a ` wrong ' choice of period
. on the other hand string theory has considered these extremal holes from a microscopic viewpoint , and used an indirect argument ( going to weak coupling ) to count the states of the system .
the entropy agrees exactly with the area entropy @xmath133 @xcite .
so the question arises : can we find a way to reconcile these two results on the entropy of extremal holes ?
we do not know how to construct all the capped geometries for the 3-charge system , but there are several pieces of evidence for the fact that the ` cap ' structure of 3-charge and 4-charge systems is analogous to that in the 2-charge case . in @xcite
capped geometries were constructed for a special family of states , and the depth of the throat upto the cap gave an energy gap that exactly agrees with the gap in the cft .
we can ` squeeze ' the 3-charge hole by compactifying a transverse direction ; this gives a 3-charge extremal system in 3 + 1 noncompact dimensions .
this system is completely analogous to the 2-charge system in 4 + 1 dimensions ; the throat narrows down as we go to @xmath134 .
an indirect argument in @xcite showed that for this system of 3-charges in 3 + 1 dimensions the throat depths for the generic state were consistent with the energy gaps in the cft .
large families of capped geometries have been found in @xcite , and the maximal throat depth for systems with @xmath135 symmetry was related to the cft energy gap in @xcite .
the metric the naive 3-charge geometry has the form ds^2=(1+qr^2)^-1 [ -dt^2+dy^2]+(1+qr^2)[dr^2+r^2d_3
^ 2]+dy^2+dz_adz_a [ naive3 ] this naive geometry is pictured in fig.[first](a ) , while the fuzzball picture of ` caps ' is fig.[first](c ) .
we now see a possible reconcilation between the two results mentioned above for the entropy of extremal holes .
the classical calculation which obtains @xmath136 considers a completely uniform infinite throat , and indeed in the fuzzball picture there are no states at the end of the infinite throat .
the string calculation , on the other hand , started with small coupling @xmath17 , where the string bound state was a small quantum object .
as we increase @xmath17 , the geometry picks up a deeper and deeper throat , but at no point does the throat become strictly infinite . to catch the states that have the entropy @xmath3 we must stop at a large but finite distance down the throat , where the generic caps appear . for each string state
the geometry at the dashed boundary of fig.[first](c ) has some distortion from spherical symmetry .
if we insisit on complete spherical symmetry we get the infinite throat , but this does not correspond to any state in the fuzzball picture , in agreement with the classical result .
we have looked at the ` size ' of the fuzzball for a sub - ensemble of states of the 2-charge system , and noted that its surface area satisfies a bekenstein type relation with respect to its entropy . at this point
we should check that quantum corrections do not make our observation meaningless ; i.e. , we should look at the sources of corrections to the classical geometries that we have used and check that their effects do not modify the essential properties of our states .
the generic state is of course not well described by a classical solution it is a quantum ` fuzzball ' .
there are two sources of quantum effects : \(a ) _ low occupation numbers @xmath137:_the classical solutions ( [ qsix ] ) were generated by starting with a classical profile @xmath22 of the string .
fourier transforming the @xmath138 in @xmath26 we arrive at a discrete set of harmonics , each of which is quantized as a harmonic oscillator . , and since the states are bps it will also give us the count at strong coupling .
for a direct computation of the phase space structure at strong coupling , see @xcite . ]
consider the oscillator for a fourier mode @xmath43 .
if the occupation number @xmath139 of this mode is @xmath140 , then we get the same essential physics whether we look at energy eigenstates @xmath141 or coherent states |>=e^-||^22e^a_k^|0 , ||^2= m but in the generic state the oscillators have occupation numbers @xmath142 ( eq.([unity ] ) ) .
so coherent states are not a good approximation , and all we have is a collection of eigenstates with low occupaption numbers for a large number of different harmonics @xmath43 .
it is not hard to see that the transverse _ size _ of the state depends on the mean harmonic number @xmath56 , while the _ fluctuations _ depend on the mean occupaption number per harmonic @xmath143 .
so the generic state has fluctuations of order unity , but we can estimate its size by using a classical state where all the energy is placed in a one harmonic of order @xmath56 .
\(b ) _ string 1-loop corrections:_we can have stringy effects that correct the state .
the leading terms of this kind give @xmath144 terms in the low energy effective action .
in @xcite it was noted that this term arises at one - string - loop order where an ns1 winding around the @xmath8 propagates around the loop . for the case of @xmath145 k3 compactifications
we get an effective @xmath146 term in the noncompact directions , and this becomes order unity near the boundary of our generic state .
it was shown in @xcite that this term remains bounded even though the @xmath8 shrinks to zero size ; this happens because the @xmath8 is a contractible loop in the full ` capped ' geometry . in the case
@xmath147 , we get just an @xmath144 term in the noncompact directions .
as we will see , the ratio of this @xmath144 term to the leading @xmath148 term is small , and we can explore a large range of the parameter @xmath11 in defining our sub - ensemble without the @xmath144 corrections becoming significant at the boundary . as noted in @xcite , the @xmath144 terms in the effective action relevant to our system arise from string 1-loop effects , where a ns1 winding around the @xmath66 circle propagates around the loop .
let us write this term for iib string theory , in the string frame , in the 8 + 1 dimensional space obtained by reduction on the @xmath8 along @xmath66 .
we get ( the notation is explained in @xcite ) s_1-loop=18 g_n^(10)g^2 ^292 ^ 6d^9 x [ r(t_8t_8 + 14_9_9)r_s^4 + 1r(t_8t_8 - 14_9_9)r_s^4 ] here @xmath149 is the radius of the @xmath8 in the direction @xmath66 .
this circle becomes small at the fuzzball boundary , so the term in @xmath150 is the one of interest to us .
the leading order term in the action is s=116g_n^(10)d^9 x ( 2r ) e^-2r_s assuming @xmath151 , the ratio of the relevant part of the 1-loop correction and the leading term is ~g^2r^2 r_s^3 [ pthree ] now let us consider the d1-d5 system and compute the value of @xmath152 at the boundary of the generic fuzzball .
this boundary is at @xmath125 ( eq . ( [ bvalue ] ) ) , and we will use the naive metric ( [ naived ] ) to compute the factors in @xmath152 . the length of the compact @xmath66 circle was computed in ( [ length1 ] ) , so we have r^2~r^2 b^2(q_1q_5)^12 [ pone ] the geometry ( [ naived ] ) is locally @xmath40 in the throat , so the curvature scale can be read off from the radius of the @xmath68 .
we have r_s~1(q_1q_5)^12 [ ptwo ] substituting ( [ pone]),([ptwo ] ) in ( [ pthree ] ) , we find ~v^2g^2 1n_1n_5 [ pfour ] to study the black hole properties of the 2-charge system we keep the moduli @xmath153 fixed and take the limit @xmath154 .
we observe that ~1n_1n_5
< < 1 so the 1-loop terms do not give a significant correction at the boundary of the fuzzball region . now let us look at the boundary of the sub - ensemble with mean harmonic order @xmath155 ( eq.([knew ] ) ) .
from ( [ bvaluenew ] ) we see that the value of @xmath73 is now smaller by a factor @xmath11 . calling @xmath156 the new radius of the @xmath66 circle , we have r_m^2=1m^2 r^2 while @xmath157 remains unchanged ( we still have the same geometry @xmath40 ) .
thus we find ~m^2v^2g^2 1n_1n_5~m^21n_1n_5 where in the last step we have again ignored the moduli .
we therefore see that in computing the area of the boundary of the cap region we can ignore the 1-loop effects as long as we take 1<<m<<(n_1n_5)^12 this allows a large range for @xmath11 , and thus allows us to consider a large set of sub - ensembles . in principle
this is what we needed to check ; if the @xmath144 corrections do not significantly affect the region at the boundary , then the area of this boundary has been correctly estimated .
we do not actually need to know the details of the state inside the fuzzball .
but since we have set up the tools to estimate length and curvature scales , we try to obtain a rough picture of the fuzzball interior by extending these estimates
. as noted at the start of this section , there are two kinds of quantum effects that we must consider .
the quantum fluctuations of type ( a ) are always there for the generic state , and gives quantum fluctuations of order unity , but this is a simple effect following from the basic quantum mechanics of harmonic oscillator wavefunctions .
we would thus like to note it and move on to exploring effect ( b ) . to do this
we assume that the occupation numbers of harmonics have been chosen to be somewhat larger than unity so that the quantum fluctuations of type ( a ) are suppressed .
given this assumption , we would like to get a general picture of the geometry in the interior of the boundary , and then see what the corrections of type ( b ) do .
let us work in the d1-d5 frame . for simplicity we will set all moduli to be order unity r~1 , v~1 , g~1 [ choice ] thus string length , 10-d planck length and 5-d planck length are all of the same order , and
this length has been set to unity l_p~l_s~1 we let the charges be equal n_1=n_5=n1 we have q_1~q_5~n note that the coordinate size of the generic state is given by |x|=b~1 using the metric ( [ qsix ] ) , we observe the following \(i ) the length of the circle @xmath8 at @xmath158 is l_y~(b^4)^14~n^-12 [ ll ] note that this is small ; its smallness is reposnible for the fact that @xmath144 terms can become significant for k3 compactifications @xcite .
\(ii ) the proper length of the diameter of the region in @xmath69 space enclosed by the boundary is d~(q_1q_5b^4)^14 b~n^12 [ dvalue ] note that the size of our ` horizon ' is therefore much larger than string or planck length , and thus the 2-charge system should be thought of as a good model for black holes .
\(iii ) now consider the picture of the ` ball ' drawn in fig.[intro](a ) .
the path of the ns1 traces out a convoluted curve inside this ball , and we wish to understand some properties of this curve . for example , we would like to know the typical distance between one segment of this curve and the closest neighboring segment .
we do this as follows .
look at the top sketch in fig.[intro](a ) .
we know that the generic state has wavenumber @xmath159 .
thus there are @xmath160 oscillations of this curve on the string .
each oscillation , roughly speaking , makes the string traverse once across the ball .
we draw the 4-dimensional space @xmath69 in fig.[ball ] .
let us make a 3-dimensional section through the center of this 4-d region
. there will be @xmath160 intersections of the curve with this plane . since the proper diameter of this ball is @xmath161 ( eq . ( [ dvalue ] ) )
, these intersections occur in a 3-dimensional area a_3~n^32 thus the density of intersection points in this plane is ~na_3~n^-12 and the typical proper distance between points on this plane will be d~1 ^ 13~n^16 [ dd ] thus we see that the strands of the curve do not come within planck or string distance of each other , but stay quite far from each other in generic configurations .
\(iv ) let us note the ` thickness ' of the curve traced out by the string .
we are working in the d1-d5 frame .
a shown in @xcite , each point on the curve is not a singularity but the center of a kk monopole .
the curvature length scale of this monopole is set by the length of the fibred circle , which is the @xmath66 circle @xmath8 , and this has length @xmath162 ( eq .
( [ ll ] ) ) .
this is very small ( much less than planck length ) .
this situation is not a surprise , since it is exactly what happens with any brane in string theory : there can be a high curvature in a planck region around the brane , and this is just included in the description of the brane .
the kk monopole tube is dualizable ( locally ) to any other brane in string theory , and so we will leave a planck sized tube around our curve as a region where we will not trust the classical metric .
\(v ) now let us look at the harmonic functions @xmath163 that give the metric ( [ qsix ] ) .
all these functions have a similar structure , so let us look at @xmath164 .
first let us go close to a point on the curve @xmath22 , say at a point @xmath165 .
near this point , the curve looks like a straight line ; let @xmath166 be the coordinate distance from this line in the coordinates @xmath69 that we have used on the noncompact space for the d1-d5 system . if we are sufficiently close to the curve @xmath22 then we can replace the curve by the line , and we find .
but with our choice of moduli ( [ choice ] ) , we have @xmath167 , @xmath168 , @xmath169 , and from @xmath170 we know that @xmath171 .
thus we get @xmath172 . ]
h_1~1 [ rho ] now consider a point , like point a in fig.[ball](a ) , which is _ not _ close to any point on the curve @xmath22 .
there is a contribution to @xmath164 from segments of @xmath22 that are close to the point a , and a contribution from segments further away .
the closer segments are fewer , but for them @xmath173 is larger ; the further segments are more numerous , but @xmath173 is smaller .
which contribution is more important ?
a simple estimate tells us that the ` far away ' contribution is larger , but note that it also changes smoothly with position , since it comes from strands that are far away from the point a. let us call it @xmath174 .
the ` nearby ' contribution is smaller , but it fluctuates on the scale of the distance between neightbouring strands , which we have seen is @xmath175 .
we call it @xmath176 .
we find h_1&=&h_1^smooth+h_1^fluch_1^smooth&~ & nh_1^fluc&~&~n^13 ( to estimate @xmath174 we note that @xmath177 and @xmath178 in ( [ functionsqq ] ) . to estimate @xmath176
we have noted that the _ coordinate _ distance @xmath166 from point a to the nearest strand is the proper distance ( [ dd ] ) divided by @xmath179 ; here the factor @xmath179 is the ratio between coordinate and proper distances in @xmath69 space .
we then substituted this @xmath166 in ( [ rho ] ) . ) the function @xmath180 has the same behavior , and in @xmath118 we find that the smooth part cancels to zero ( since the orientation of the vector @xmath181 changes randomly along the curve @xmath22 ) while the fluctuating part has the same order as @xmath176 .
since the functions @xmath163 appear as coefficients in the metric ( [ qsix ] ) , we find that the metric in the space @xmath69 ( schematically called @xmath17 ) itself has a smooth part and a fluctuating part g = g^smooth+g^fluc with h~n^-23 [ hhh ] \(vi ) finally , we use the length scales found above to examine the @xmath144 effects .
the smooth part of the metric @xmath182 varies over a distance of the order of the radius @xmath179 of the ball .
thus it gives a curvature @xmath183 .
thus for @xmath152 we have ~n^-3n^-1~1n^2 < < 1 where we have used ( [ ll ] ) for @xmath184 . the fluctuating part varies on a smaller distance scale @xmath185 , but @xmath186 is also small .
thus we get a curvature r^fluc_s~hd^2~n^-1 which again gives @xmath187 . to summarize
, we get the picture sketched in fig.[ball ] .
all lengths are measured in units of planck length .
we have a ball of diameter @xmath161 .
in this balls a curve winds around ; the quantum ` thickness ' of this curve is unity .
the strands of this curve do not come too close to each other ; the typical distance to the nearest neighbor is @xmath175 .
the @xmath144 corrections are small .
the ordinary quantum fluctuations of type ( a ) are always present , and these will cause the curve to fluctuate , and be in a general quantum wavefunction . 0.7 truein . in the generic state
the location of the thick curve has quantum fluctuations that are large but these have been suppressed ( by choosing high occupation numbers of modes ) to enable an estimate of the string 1-loop effects .
, title="fig:",width=336 ]
we have looked at some subsets of states of the 2-charge extremal system , and found that in each case the boundary of the ` fuzzball ' satisfies a bekenstein type relation with the entropy inside the ball .
this suggests that the system requires a minimum sized region to contain its phase space , and black holes ` pack states ' into these minimal regions .
let us note some ways in which phase space considerations can impact the puzzles arising with black holes . in the classical treatment of a black hole we do not see the states of the hole ( ` black holes have no hair ' ) .
thus we miss an essential feature of the hole , since the entropy is actually very large .
extremal holes may be the easiest to understand .
we can start with the brane bound state at weak coupling @xmath17 ; this state has a large degeneracy .
since the states are bps , we can follow them to strong coupling .
the geometry develops a throat , and the throat gets deeper with increasing @xmath17 , but by continuity in @xmath17 we do not expect at any given @xmath17 to have an infinite throat with a horizon at the end .
the ` caps ' at the end of the throat have a structure that reflects the particular microstate . in the 2-charge d1-d5 system
the ` true charges ' d1,d5 give rise to a ` dipole ring ' which takes the form of a kk monopole times a @xmath8 .
different shapes of this ring give different classical geometries .
quantum fluctuations ( which we reviewed in this paper ) make the generic state a ` fuzzball ' with the same size as observed in the classical geometries . for three and four charge systems
the dipole charges give more complex structures to the cap , and large families of classical solutions have been found starting from these dipole charges @xcite . in the 4-charge solutions the dipoles are points in the noncompact space .
as we go to more and more generic states , the number of these dipole charges can be expected to grow , and we will have a vast collection of ` branes at angles ' in the ` cap ' region @xcite .
we can understand all d - branes as arising from tachyon condensation in other sets of branes and anti - branes ; this is the ` k - theory ' picture of d - branes .
thus we can visualize the generic ` cap ' state as a wavefunction in this very complex potential made from an arbitrary number of branes .
if the tachyon condenses to its minimum in a given charge sector then we see the bps dipole branes observed in the classical solutions @xcite , but more generally the system will just have a complicated wavefunction in this multi - dimensional tachyon potential .
the entropy of non - bps holes can be understood in terms of brane - antibrane pairs @xcite , so for these holes we expect a similar potential with a large number of local extrema ; the slow decay by hawking radiation would then correspond to the system drifting to its true ground state .
if this picture of black hole states is correct , then we must understand the entropy of the hole in terms of the number of metastable states possible in the tachyon potential arising from an arbitrary number of brane pairs .
put another way , if we are given an energy @xmath188 , then the maximal entropy is not obtained by converting this energy into massless quanta , but rather by using it to create brane and anti - branes , which is a system with a large phase space , and gives a high degeneracy of states .
the traditional puzzle with black holes has been the following .
we have a given classical geometry .
there seem to be no large corrections at the horizon .
why should string theory change anything ?
let us address this question in the context of extremal holes .
the classical geometry has an infinite throat .
nothing seems to be wrong with this geometry by itself .
but the recent work on fuzzballs shows us that there are a large number of _ other _ states with the same mass and charge , where the throat behaves differently after a certain depth .
why are these states relevant ? if we quantize the system , we have one state per unit cell in phase space . but to find this unit cell , we first have to understand all the directions of this phase space . to illustrate this point schematically , consider a 1-dimensional space ( drawn as the horizontal line in fig.[phase ] ) .
since this line is infinite , it would appear that there will be an infinite number of unit cells along its length . but
suppose there was another direction @xmath189 to this phase space , and that the volume of phase space ( available at a given energy ) narrows down to the right as shown in the figure .
now we see that beyond some point @xmath190 there is very little phase space left , and this available phase space might be able to hold only a finite number of states .
axis denotes one direction of a phase space ; this would appear to have an infinite number of unit cells beyond the point @xmath190 .
the @xmath189 axis denotes another direction of phase space ; since this narrows to the right there are only a finite number of states to the right of @xmath190 in the full phase space.,width=384 ] it may be possible to see such a situation in the 2-charge extremal system .
first let us split our d1-d5 system with charges @xmath29 into two smaller systems , with charges @xmath191 and @xmath192 respectively .
the relative motion of these two objects is described by a moduli space metric , which we can find from @xcite by restricting their 3-charge answer to the case where only two charges are present .
we find ( for small separations @xmath193 ) ds^2=(n^(1)_1n^(2)_5+n^(1)_5n^(2)_1)r^2(dr^2+r^2d_3 ^ 2)= ( n^(1)_1n^(2)_5+n^(1)_5n^(2)_1)[dr^2r^2+d_3 ^ 2 ] [ je ] we see that the angular part of this moduli space has a constant volume , while the radial part is an infinite line .
( we set @xmath194 and find the moduli space metric @xmath195 for @xmath196 . )
this suggests an infinite number of quantum states .
we have a single throat that branches into two throats after a great depth , so we are quantizing some degrees of freedom of the geometry of the 2-charge throat . ]
but we know that there are no throats longer than a maximal depth in the correct quantization of the 2-charge system , and there are a finite number of states overall . in the cft
dual , the longest throat corresponds to a single ` component string ' of length @xmath197 @xcite . in the ns1-p duality frame
, the longest throat corresponds to having a single vibration in the harmonic @xmath198 carrying all the momentum charge .
we can not go to higher harmonics of vibration , since then we will need a fractional number of quanta to carry the given momentum charge
. this fractional number of quanta can be thought of as an excitation that was allotted a fractional unit of phase space to exist ; such an excitation is not possible .
the above considerations were arrived at using the cft dual , but the phase space of the d1-d5 system can also be quantized directly in the gravity description , and the same measure on phase space is obtained @xcite .
thus we find that when we quantize the full phase space of the system , rather than just one direction like in ( [ je ] ) , then all throats longer than a given depth fall into a single wavefunction , and it does not make sense to explore the throat to arbitrary depths in the manner suggested by the classical metric .
we hope to return to these issues in the future .
i am grateful to stefano giusto for help with many aspects of this paper .
i would like to thank borundev chowdhury and jeremy michelson for helpful comments .
i would also like to thank iosif bena , j. de boer , eric gimon , tom levi and nick warner for sharing their thoughts on related projects .
this work was supported in part by doe grant de - fg02 - 91er-40690 .
j. d. bekenstein , phys .
d * 7 * , 2333 ( 1973 ) .
a. strominger and c. vafa , phys .
b * 379 * , 99 ( 1996 ) [ arxiv : hep - th/9601029 ] . c. g. .
callan and j. m. maldacena , nucl .
b * 472 * , 591 ( 1996 ) [ arxiv : hep - th/9602043 ] .
s. d. mathur , nucl .
b * 529 * , 295 ( 1998 ) [ arxiv : hep - th/9706151 ] .
o. lunin and s. d. mathur , nucl .
b * 623 * , 342 ( 2002 ) [ arxiv : hep - th/0109154 ] .
s. giusto , s. d. mathur and a. saxena , nucl .
b * 701 * , 357 ( 2004 ) [ arxiv : hep - th/0405017 ] ; s. giusto , s. d. mathur and a. saxena , arxiv : hep - th/0406103 ; o. lunin , jhep * 0404 * , 054 ( 2004 ) [ arxiv : hep - th/0404006 ] . s. giusto and s. d. mathur , arxiv : hep - th/0412133 .
i. bena and n. p. warner , adv .
* 9 * , 667 ( 2005 ) [ arxiv : hep - th/0408106 ] ; i. bena and n. p. warner , phys .
d * 74 * , 066001 ( 2006 ) [ arxiv : hep - th/0505166 ] ; i. bena , c. w. wang and n. p. warner , arxiv : hep - th/0604110 ; i. bena and n. p. warner , arxiv : hep - th/0701216 .
i. bena , c. w. wang and n. p. warner , jhep * 0611 * , 042 ( 2006 ) [ arxiv : hep - th/0608217 ] ; p. berglund , e. g. gimon and t. s. levi , jhep * 0606 * , 007 ( 2006 ) [ arxiv : hep - th/0505167 ] .
v. balasubramanian , e. g. gimon and t. s. levi , arxiv : hep - th/0606118 .
a. saxena , g. potvin , s. giusto and a. w. peet , jhep * 0604 * , 010 ( 2006 ) [ arxiv : hep - th/0509214 ] ; j. ford , s. giusto and a. saxena , arxiv : hep - th/0612227 .
o. lunin and s. d. mathur , phys .
* 88 * , 211303 ( 2002 ) [ arxiv : hep - th/0202072 ] .
s. d. mathur , fortsch .
phys . * 53 * , 793 ( 2005 ) [ arxiv : hep - th/0502050 ] .
o. lunin , j. maldacena and l. maoz , arxiv : hep - th/0212210 .
m. taylor , jhep * 0603 * , 009 ( 2006 ) [ arxiv : hep - th/0507223 ] .
l. f. alday , j. de boer and i. messamah , jhep * 0612 * , 063 ( 2006 ) [ arxiv : hep - th/0607222 ] .
a. sen , nucl .
b * 440 * , 421 ( 1995 ) [ arxiv : hep - th/9411187 ] ; a. sen , mod .
lett . a * 10 * , 2081 ( 1995 ) [ arxiv : hep - th/9504147 ] .
o. lunin and s. d. mathur , nucl .
b * 610 * , 49 ( 2001 ) , hep - th/0105136 .
i. bena and p. kraus , jhep * 0412 * , 070 ( 2004 ) [ arxiv : hep - th/0408186 ] ; l. f. alday , j. de boer and i. messamah , nucl .
b * 746 * , 29 ( 2006 ) [ arxiv : hep - th/0511246 ] .
f. denef and g. w. moore , arxiv:0705.2564 [ hep - th ] .
m. srednicki , phys .
* 71 * , 666 ( 1993 ) [ arxiv : hep - th/9303048 ] .
l. susskind , j. math .
phys . * 36 * , 6377 ( 1995 ) [ arxiv : hep - th/9409089 ] ; j. m. maldacena , adv .
* 2 * , 231 ( 1998 ) [ int .
. phys . * 38 * , 1113 ( 1999 ) ] [ arxiv : hep - th/9711200 ] . j. d. brown and m. henneaux , commun . math .
phys . * 104 * , 207 ( 1986 ) ; s. carlip , class .
quant . grav .
* 22 * , 1303 ( 2005 ) [ arxiv : hep - th/0408123 ] .
r. m. wald , phys .
d * 48 * , 3427 ( 1993 ) [ arxiv : gr - qc/9307038 ]
. v. s. rychkov , jhep * 0601 * , 063 ( 2006 ) [ arxiv : hep - th/0512053 ] .
i. kanitscheider , k. skenderis and m. taylor , arxiv:0704.0690 [ hep - th ] .
j. preskill , p. schwarz , a. d. shapere , s. trivedi and f. wilczek , mod .
lett . a * 6 * , 2353 ( 1991 ) ; g. w. gibbons and r. e. kallosh , phys . rev .
d * 51 * , 2839 ( 1995 ) [ arxiv : hep - th/9407118 ] .
s. w. hawking , g. t. horowitz and s. f. ross , phys .
d * 51 * , 4302 ( 1995 ) [ arxiv : gr - qc/9409013 ] ; c. teitelboim , phys .
d * 51 * , 4315 ( 1995 ) [ erratum - ibid .
d * 52 * , 6201 ( 1995 ) ] [ arxiv : hep - th/9410103 ] .
f. denef , jhep * 0210 * , 023 ( 2002 ) [ arxiv : hep - th/0206072 ] .
j. evslin , arxiv : hep - th/0610328 . g. t. horowitz , j. m. maldacena and a. strominger , phys . lett .
b * 383 * , 151 ( 1996 ) [ arxiv : hep - th/9603109 ] .
v. jejjala , o. madden , s. f. ross and g. titchener , phys .
d * 71 * , 124030 ( 2005 ) [ arxiv : hep - th/0504181 ] ; e. g. gimon , t. s. levi and s. f. ross , arxiv:0705.1238 [ hep - th ] . | for extremal black holes the fuzzball conjecture says that the throat of the geometry ends in a quantum ` fuzz ' , instead of being infinite in length with a horizon at the end . for the d1-d5 system
we consider a family of sub - ensembles of states , and find that in each case the boundary area of the fuzzball satisfies a bekenstein type relation with the entropy enclosed .
we suggest a relation between the ` capped throat ' structure of microstate geometries and the fact that the extremal hole was found to have zero entropy in some gravity computations .
we examine quantum corrections including string 1-loop effects and check that they do not affect our leading order computations .
epsf c black hole size and phase space volumes + + department of physics , + the ohio state university , + columbus , oh 43210 , usa + |
the nuclear rotation is one of the most typical collective motions in nuclei @xcite .
its semiclassical nature makes it possible to introduce the classical concepts like the rotational frequency and the transformation to the rotating frame , which are very useful to analyze the high - spin rotational bands @xcite and nowadays provide a standard method called the cranked shell model @xcite .
however , the nucleus is a quantum many - body system and the basis of such a semiclassical treatment of the rotational motion is the symmetry breaking caused by the deformed mean - field , see e.g. @xcite .
in fact , the rotational motion emerges as a symmetry restoring collective motion of atomic nucleus as a whole and can be described full - quantum mechanically by the angular - momentum - projection method @xcite .
although a nice rotational spectrum can be obtained by the angular - momentum - projection from the deformed mean - field state , it has been known that the level spacing of the obtained rotational spectrum tends to be larger than that of the experimental data ; i.e. the moment of inertia is quite often underestimated .
inclusion of the time - odd components into the deformed mean - field , from which the projection performed , improves this problem and it can be easily realized by the cranking procedure @xcite .
it has been demonstrated that the small cranking frequency is enough to increase the moment of inertia and the result of projection does not depend on the actual value of the frequency ; we call it `` infinitesimal cranking '' for the angular - momentum - projection @xcite .
recently we have extended the study of the rotational motion by the angular - momentum - projection method @xcite . namely the cranking procedure is combined with the projection by employing the configuration - mixing with respect to the finite rotational frequency
; we call it angular - momentum - projected multi - cranked configuration - mixing .
this method was originally proposed by peierls and thouless long time ago @xcite , but has not been taken seriously .
we have applied it to a few examples to show that it gives a reliable description of the rotational motion at high - spin states @xcite .
the angular - momentum - projected configuration - mixing with respect to a few cranked mean - field states has recently been performed also in ref .
@xcite . as for the application of the angular - momentum - projection method to the nuclear rotational motion ,
many pioneering works have been done by the projected shell model , see e.g. ref . @xcite . while the basic idea is the same , much larger but simple configurations , like the zero- , two- , four- ,
... , quasiparticle excited bands are mixed in the sense of the shell model .
we believe the multi - cranked configuration - mixing @xcite is an alternative , which incorporates a relatively small number of mean - field configurations with the help of the cranking procedure .
the main purpose of the present work is to demonstrate that the multi - cranked configuration - mixing is indeed a reliable method to describe the rotational band with the angular - momentum - projection method .
we first apply the method to the ground - state rotational bands for a number of selected nuclei in the rare earth region . at high - spin states
, it is well - known that the band crossing ( back - bending ) phenomenon between the ground - state ( g- ) band and the stockholm ( s- ) band , i.e. , the two - neutron aligned band , occurs .
therefore , we try to study the s - band in a typical nucleus @xmath0er with the same multi - cranked configuration - mixing method ; we are able to study the g- and s - bands separately without the inter - band mixing .
the cranked mean - field states are determined selfconsistently by the hartree - fock - bogoliubov ( hfb ) method for given rotational frequencies employing the finite - range gogny interaction @xcite with the d1s parameter set @xcite .
after briefly explaining the theoretical framework in sec .
[ sec : multi ] , we show the results of calculations in sec . [
sec : results ] .
the conclusion is drawn in sec .
[ sec : concls ]
our basic approach to study the high - spin states of the nuclear collective rotation is the angular - momentum - projected configuration - mixing , or the projected generator coordinate method ( gcm ) , where the cranking frequency @xmath1 is employed as a generator coordinate .
it was first proposed by peierls - thouless @xcite , and the wave function is calculated by @xmath2 where the operator @xmath3 is the angular momentum projector , and the mean - field wave function , @xmath4 , is obtained by the selfconsistent cranking procedure with the cranked hamiltonian , @xmath5 , @xmath6 in the present work , the ground - state mean - field states are axially deformed and the cranking axis is chosen to be the @xmath7-axis perpendicular to the symmetry axis ( @xmath8-axis ) . practically we discretize the generator coordinate , i.e. , the cranking frequency , as ( @xmath9 ; @xmath10 ) in eq .
( [ eq : ptanz ] ) , @xmath11 and solve the configuration - mixing amplitude , @xmath12 , with the so - called hill - wheeler equation , @xmath13 where the hamiltonian and norm kernels are defined as usual , @xmath14 we do not perform the number projection in the present work , and treat the number conservation approximately by replacing @xmath15 , where @xmath16 and @xmath17 are the neutron and proton numbers to be fixed .
as for the neutron and proton chemical potentials @xmath18 and @xmath19 we use those obtained for the hfb ground - state .
we have recently developed an efficient method for the angular - momentum - projection and the configuration - mixing @xcite .
this method is fully utilized also in the present work . to solve the hfb equation and to perform the projection calculation
the harmonic oscillator basis expansion is employed .
more details of our theoretical framework can be found in refs .
as it is mentioned we employ the gogny force with the d1s parameter set @xcite as an effective interaction .
therefore there is no ambiguity for the hamiltonian .
the hfb equation is solved in the space generated by the isotropic harmonic oscillator potential with the frequency @xmath20 mev .
the size of the space is controlled by the oscillator quantum number @xmath21 ; all the basis states satisfying @xmath22 are included .
we use @xmath23 for the following systematic calculations of the rotational spectra .
the main target of the present work is the most basic rotational band , i.e. , the g - band .
therefore , we selected typical deformed nuclei in the rare earth region , i.e. , three isotopes with the neutron number in @xmath24100 in each gd , dy , er , and yb nuclide , as they are tabulated in table [ tab : mf_rare ] . in this table
we show the nuclear radii , the deformation parameters , and the average pairing gaps obtained by the hfb calculations in the ground - states of these selected nuclei .
the ground - states are axially symmetric in all nuclei and the deformation parameter @xmath25 is defined by @xcite @xmath26 where the @xmath27-pole moment @xmath28 and the radius @xmath29 are calculated by the expectation value with respect to the hfb state , @xmath30^{1/2}. \label{eq : qmom}\ ] ] the average pairing gap is defined by @xmath31 \left[\sum_{a>0}\kappa^*_{a\tilde{a}}\right]^{-1},\qquad \delta_{ab}=
\sum_{c > d}\bar{v}_{ab , cd}\,\kappa_{cd } , \label{eq : avgap}\ ] ] where the quantities @xmath32 and @xmath33 are the anti - symmetrized matrix element of the two - body interaction and the abnormal density matrix ( pairing tensor ) , respectively @xcite , and @xmath34 means the time - reversal conjugate state of @xmath35 .
here we corrected the misprinted expression of denominator in the definition of @xmath36 in refs.@xcite .
the even - odd mass differences calculated by the 4th - order difference formula based on the 2003 mass table @xcite are also included in table [ tab : mf_rare ] . [ cols="^,<,^,<,^,<,^,<,^,^,<,^,^,<,^ " , ] the result of the deformation parameter @xmath37 roughly corresponds to experimental data deduced by the measured @xmath38 values @xcite , but the calculated @xmath37 are slightly smaller .
this is merely due to the differences in the definitions of @xmath37 in ref .
@xcite and in the present work .
the calculated radius @xmath29 is 2.42.9% larger than the empirical value @xmath39 [ fm ] , which is mostly due to the effect of deformation . in these stable nuclei
the difference of the radius from the empirical value is not so large , but for the unstable nuclei the difference is expected to be larger .
therefore , it is important to measure the nuclear radius to reliably extract the deformation parameters @xcite . as for the average pairing gaps selfconsistently calculated values
are smaller than the even - odd mass differences in most cases . especially the neutron pairing gap in @xmath40gd is only 65% of the even - odd mass difference , and in @xmath41yb the calculated average gaps for both the neutron and proton are about 27% smaller .
on the other hand , for @xmath42gd , @xmath43dy and @xmath0dy , the calculated gaps relatively well correspond to the even - odd mass differences , and their differences are less than 14% . as it is discussed in the following ,
the agreement of the calculated gaps with the even - odd mass differences is crucial to reproduce the moments of inertia of the g - band .
if the deformed superconducting state is obtained for the ground - state , we generate the cranked hfb states in eq .
( [ eq : crank ] ) for a given set of rotational frequencies ( @xmath9 ; @xmath10 ) .
it was demonstrated @xcite that the result of configuration - mixing does not depend on the choice of a set of frequencies , if the number of frequencies , @xmath44 , is five .
therefore , we take @xmath45 and choose them ( almost ) equidistantly .
since the ground - state rotational band is studied , only the frequencies before the g - s crossing should be selected ; we choose @xmath46 mev as in ref .
@xcite , for most of nuclei in table [ tab : mf_rare ] . however , it is known that there is no sharp g - s crossing observed in n=98 isotopes , so that we choose @xmath47 mev for @xmath0dy and @xmath48yb .
once the cranked hfb states are prepared we perform the angular - momentum - projected multi - cranked configuration - mixing , see eq .
( [ eq : proj ] ) , to obtain the rotational spectrum .
in order for the efficient calculation of projection , the cut - off of the quasiparticle basis is employed @xcite , namely the canonical basis states of the hfb wave function whose occupation numbers are larger than @xmath49 are only retained . as for the integration mesh points for the euler angles ( @xmath50 ) in the angular - momentum - projector , we take @xmath51 and @xmath52 . for the present systematic calculation of the ground - state band , @xmath53 and @xmath54 are chosen ; the smaller value of @xmath55 is enough because the hfb wave function is nearly axially symmetric ( @xmath56 mixing is mainly induced by the cranking procedure ) . in the configuration - mixing in eq .
( [ eq : proj ] ) , the superposition of the states with respect to @xmath57 for given @xmath58 is overcomplete and there are vanishingly small norm states , which causes numerical problems @xcite .
therefore , the norm cut - off should be done ; namely , the eigenvalues of the norm kernel are first calculated and the small norm states should be excluded when solving the hill - wheeler eq .
( [ eq : hw ] ) .
the value for the norm cut - off is better to be as small as possible not to miss important contributions .
we start from the value @xmath59 and increase it to avoid the numerical problems in each case .
the actual values used in the following calculation are also denoted in table [ tab : mf_rare ] ; they are in the range , @xmath60 .
figures [ fig : eir1 ] and [ fig : eir2 ] show the resultant rotational spectra of the configuration - mixing for the ground - state bands in nuclei in table [ tab : mf_rare ] .
the results of the simple projection from one cranked hfb state with small frequency @xmath61 mev , i.e. , those of the infinitesimal cranking , are also included .
the experimental data are taken from the table of isotope homepage @xcite . to show the detail we subtract the reference rotational energy , @xmath62 mev , in each spectrum . by comparing the results of the configuration - mixing and of
the simple projection from one infinitesimally cranked hfb state in these figures , it is clear that the infinitesimal cranking gives a good description of the low - spin states .
however , the deviations become non - negligible quickly at higher spins , @xmath63 . with the multi - cranked configuration - mixing the energy gain at @xmath64 from the simple projection
is about @xmath65 mev , and therefore the effect of configuration - mixing is crucial for the description of the high - spin states . in comparison with the experimental spectra ,
the results of the configuration - mixing reproduce the bending - down behaviors of the data at higher spins , which reflects the increase of moments of inertia .
in contrast , the spectra of the simple projection with the infinitesimal cranking do not change or even increase as spin increases , which means that the moments of inertia obtained by the simple projection are rather constant .
the deviation of the calculated spectrum from the measured spectrum is rather large in @xmath40gd , @xmath66dy , @xmath42er , and @xmath41yb , for which the average pairing gaps are considerably smaller than the even - odd mass differences , especially for neutron , see table [ tab : mf_rare ] . on the other hand ,
the agreements are almost perfect for @xmath42gd , @xmath43dy , @xmath0dy , and @xmath0yb , in which the calculated average pairing gaps for both neutron and proton well correspond to the even - odd mass differences .
thus the reproduction of the pairing properties is very important to achieve a good description of the ground - state rotational band , which is a rather well - known fact .
it should be emphasized that the agreements for other nuclei are rather satisfactory considering the fact that we have no room for adjustment in the present calculations .
the relative difference between the isotopes , e.g. , between @xmath43er and @xmath0er , or between @xmath0yb and @xmath48yb , is also reproduced .
to study the rotational property in more detail , we show the first ( or kinematic ) moment of inertia in figs .
[ fig : emo1 ] and [ fig : emo2 ] , which is defined by @xmath67 as a reference , the result of the cranked hfb calculation is also included in these figures , which is calculated by @xmath68 and is plotted as a function of @xmath69 we do not try to search the minimum energy at the fixed spin value in the cranked hfb calculation , and therefore the back - bending behavior of the moment of inertia is not obtained .
it should be noticed that the cranked hfb inertia in eq .
( [ eq : j1chfb ] ) at high - spin states after the alignment of two quasineutrons should be considered to be that of the s - band and the inertia is unphysical in the band crossing region .
the irregularities seen in the experimental moment of inertia are due to the effect of the g - s band crossing .
it can be seen in figs .
[ fig : emo1 ] and [ fig : emo2 ] that the moments of inertia for the ground - state band increase gradually as functions of spin .
this behavior is quite nicely reproduced by the multi - cranked configuration - mixing calculations .
the values of moment of inertia are considerably overestimated at low - spins in @xmath40gd , @xmath66dy , @xmath42er , and @xmath41yb ; again this is mainly because the calculated pairing gaps are markedly smaller than the even - odd mass differences . in these nuclei
the amounts of increase for the moment of inertia are also much smaller than the experimental data .
this indicates that the increase of the moment of inertia is mainly related to the reduction of the pairing correlations at higher spins .
the agreements between the calculated and measured inertias are excellent for @xmath42gd , @xmath43dy , @xmath0dy , and @xmath0yb in the whole spin range shown in the figures . for the other cases , @xmath66gd , @xmath43er , @xmath0er , and @xmath48yb , the deviations of the calculated moments of inertia from the experimental data are less than 20% , which is quite non trivial .
the different spin - dependence observed between the isotopes , e.g. , between @xmath43er and @xmath0er , is also nicely reproduced by the configuration - mixing calculations .
the inertia calculated by the cranked hfb at low - spin is slightly larger than the result of configuration - mixing ; the large increase of the cranked hfb inertia is caused by the effect of the two - neutron alignment mentioned above .
it should be emphasized that the moment of inertia obtained by the simple projection from the non - cranked hfb state ( not shown in the present work ) is about 3040% smaller than the result of the infinitesimal cranking @xcite .
the inclusion of the time - odd components of the wave functions with @xmath70 and the subsequent @xmath56-mixing is very important to reproduce the correct magnitude of the moments of inertia . in figures [ fig : emm1 ] and
[ fig : emm2 ] we show how the moments of inertia is changed by the multi - cranked configuration - mixing .
namely , the inertias calculated by the simple projection from one intrinsic hfb state with five frequencies are compared with the result of the configuration - mixing
. the calculated inertias with higher cranking frequencies are generally larger because of the coriolis anti - pairing effect .
however , those calculated by the simple projection from one hfb states are almost constant as spin increases or even decrease in several cases . in contrast , the resultant inertias of the configuration - mixing always increase as functions of spin in accordance with experimental data . in this way
the configuration - mixing is important to obtain gradually increasing behavior , which is general for the g - band in the rare earth region .
note , however , that the amount of increase is quite different in each nucleus ; there is a trend that if the difference between the inertias calculated with different cranking frequencies at high - spin is large , then the amount of increase is larger .
it is interesting to note that the behaviors of calculated inertias by the simple projection with finite frequencies are rather different in each nucleus .
for example , all five results of the simple projection are similar in @xmath0er , while they are considerably different in @xmath41yb .
however , the results of the configuration - mixing make the behaviors of moment of inertia in all nuclei rather similar , i.e. , gradually increasing as functions of spin .
it is worthwhile mentioning that the result of the infinitesimal cranking , i.e. , the calculated inertia with @xmath61 mev coincides with that of the configuration - mixing at low - spins in most cases .
this clearly shows that the infinitesimal cranking is enough for a good description of the rotational band at low - spin states , although the configuration - mixing is crucial at high - spin states . in the previous section only the g - band
is considered . however , it is well - known that a different rotational band intersects with the g - band and becomes the yrast state at higher spin values .
this band is called the s - band , in which two quasineutrons are excited to align their angular momenta to the axis of collective rotation .
the band crossing between the g- and s - bands is the origin of the back - bending phenomenon first observed in ref .
@xcite , where the rotational frequency decreases when spin increases along the yrast line .
how this alignment of the two quasineutron occurs can be nicely understand by the semiclassical cranking model , see e.g. refs .
we have shown that the g - band can be nicely described by our multi - cranked configuration - mixing method .
therefore it is natural to study the s - band with the same method .
the reference rotational energy , @xmath62 mev , is subtracted .
the result of the projected multi - cranked configuration - mixing ( mixed ) is compared with the experimental data ( exp . ) .
, width=302 ] er obtained by the simple projection from one intrinsic hfb state with five values of the cranking frequencies , @xmath710.01 , 0.05 , 0.10 , 0.15 , 0.20 mev , compared with the result of the projected configuration - mixing employing those five cranked hfb states . the energy origin is taken as the energy of the ground state of the configuration - mixing calculation . , width=302 ] er obtained by the simple projection from one intrinsic hfb state with four values of the cranking frequencies , @xmath710.25 , 0.30 , 0.35 , 0.40 mev , compared with the result of the projected configuration - mixing employing those four cranked hfb states . the energy origin is taken as the energy of the ground state of the configuration - mixing calculation . ,
width=302 ] we have already investigated the g - band in @xmath0er in our previous work @xcite with @xmath72
. therefore , we use the same model space for the s - band in this section .
the method of the calculation is the same as in ref .
@xcite except that the set of cranking frequencies is suitably chosen for the description of the s - band . by the cranked hfb calculation , the alignment of the two quasineutrons occurs at @xmath73 mev in @xmath0er with the gogny d1s parameter set , see fig . 5 of ref .
we use the four cranked hfb states with frequencies , @xmath710.25 , 0.30 , 0.35 , 0.40 mev , for the configuration - mixing calculation of the s - band .
the value of the norm cut - off is taken to be @xmath74 , with which the result is stabilized , and @xmath75 and @xmath76 are used . the resultant spectrum for the s - band
is depicted in fig .
[ fig : er164gsbr ] ; for completeness the calculated result of the g - band in ref .
@xcite is also included , which was obtained with the set of five frequencies @xmath710.01 , 0.05 , 0.10 , 0.15 , 0.20 mev , as in the previous section .
again , the reference rotational energy , @xmath62 mev , is subtracted .
as it can be seen in the figure , we have successfully obtained the band crossing between the g- and s - bands , although the crossing occurs at slightly higher spin compared with the observation ; at @xmath77 in the calculation , while between @xmath78 and 16 in the experimental data .
this is non trivial because we have no kind of adjustment in the present calculation .
it should be emphasized that the g- and s - bands are calculated independently without the inter - band mixing ; the effect of the inter - band mixing is not taken into account in the present work .
the reason why the calculated crossing is delayed is that the calculated excitation energies of the s - band are higher than the experimental data ; the excitation energy of @xmath79 state is 2.69 mev , which is about 170 kev higher than the experimentally measured one . in figs .
[ fig : er164gb5 m ] and [ fig : er164sb4 m ] , we show how the resultant spectrum is obtained by the configuration - mixing calculation for the g- and s - bands , respectively , where the five ( four ) spectra calculated by the projection from one cranked hfb state with different frequencies are depicted in addition to the result of the configuration - mixing for the g - band ( s - band ) . from fig .
[ fig : er164gb5 m ] it can be seen that the five spectra for the g - band , each of which is obtained from the one cranked state with @xmath80 mev , are rather similar , and the energy gain by the configuration - mixing is larger at higher spin , which leads to the increase of the moment of inertia as a function of spin .
this suggests that the @xmath56-mixing and configuration - mixing induced by the cranking is more effective at higher spins ; the hill - wheeler equation ( [ eq : hw ] ) should be solved even for the case of projection from a single hfb state and its dimension of increases as spin increases .
in contrast , the four spectra for the s - band obtained from the one cranked state with @xmath81 mev are rather different as is shown in fig .
[ fig : er164sb4 m ] .
non of the spectra looks like the observed one and the configuration - mixing is very important for the s - band to obtain the correct spectrum .
each spectrum in figs .
[ fig : er164sb4 m ] has a minimum energy at a finite spin value , and the spin value that gives a minimum is larger for the spectrum obtained from the cranked hfb with larger cranking frequency .
moreover , the energy gain by the configuration - mixing is considerable at lower spin , while it is much smaller at @xmath82 and the resultant configuration - mixed spectrum looks more like the envelope curve of the four spectra .
note that the aligned angular momentum of the two quasineutrons is estimated to be about 10 @xmath83 in @xmath0er .
thus , the role played by the configuration - mixing seems to be somewhat different in the g- and s - bands , and the configuration - mixing is much more crucial for the s - band than for the g - band .
er obtained by the projected configuration - mixing calculations in comparison with the experimental data . the result of the cranked hfb ( chfb ) is also included . , width=340 ] although the nice band crossing is obtained by the calculation , the agreement of the moment of inertia for the s - band is not as good as that for the g - band , which is shown in fig .
[ fig : er164moim ] . especially for the s - band , the inertia is considerably overestimated in @xmath84 , and the experimentally observed inertia decreases as spin increases , while the calculated one is almost constant in the range @xmath85 and increases afterward .
the overestimation of the moment of inertia for the s - band is mainly due to the fact that the neutron pairing vanishes by the alignment of two quasineutrons , namely , the cranked hfb states used for the configuration - mixing calculation for the s - band are in fact the unpaired states ( the slater determinants ) for neutrons .
the proton pairing is non vanishing but reduces considerably ; i.e. , the average pairing gap of proton is @xmath86 mev at @xmath87 mev and @xmath88 mev at @xmath89 mev .
this fact is reflected also in the resultant inertia of the cranked hfb calculation , which is considerably larger at @xmath84 .
because of this problem we do not discuss the higher spin part at @xmath90 in the present work .
but the calculations with employing five values of the cranking frequencies , @xmath710.250 , 0.275 , 0.300 , 0.325 , 0.350 mev . ,
width=302 ] er obtained by the configuration - mixing calculations with the two different sets of rotational frequencies , @xmath710.25 , 0.30 , 0.35 , 0.40 mev ( mixed1 ) , and @xmath710.250 , 0.275 , 0.300 , 0.325 , 0.350 mev ( mixed2 ) in comparison with the experimental data .
the scale of the ordinate is enlarged from that of fig .
[ fig : er164moim ] .
, width=302 ] in the case of the g - band the result of the configuration - mixing does not depend on the choice of a set of frequencies for the cranked hfb states @xcite .
however , it turns out that the result for the s - band is not completely independent of the choice of frequencies .
we show an example of another set of the cranking frequencies , @xmath710.250 , 0.275 , 0.300 , 0.325 , 0.350 mev in fig .
[ fig : er164sb5 m ] , where the range of the frequencies are reduced from @xmath91 $ ] to @xmath92 $ ] and they are chosen more densely in the given interval . comparing the resultant configuration - mixed spectra obtained with the two sets of cranking frequencies in figs .
[ fig : er164sb4 m ] and [ fig : er164sb5 m ] , one can see they are slightly different . to see the difference more clearly
, we show the moments of inertia for the s - band calculated with these two sets of frequencies in fig .
[ fig : er164moid ] in comparison with the experimental data .
note that the scale of the ordinate is enlarged .
the difference is non - negligible especially in the lower spin range @xmath93 .
moreover , the inertia obtained with the first set , @xmath710.25 , 0.30 , 0.35 , 0.40 mev ( mixed1 ) , is monotonically increasing as spin increases , while the one with the second set , @xmath710.250 , 0.275 , 0.300 , 0.325 , 0.350 mev , decreases first in @xmath94 and then turns to increase .
we think that the result with the second set ( mixed2 ) is more reliable in the considered spin - range because a larger number of the cranked hfb states are employed in the relevant frequency interval .
the expectation values of spin in eq .
( [ eq : iavchfb ] ) for the cranked hfb states are @xmath95 19.3 , 23.6 , 29.1 , 35.1 at @xmath71 0.25 , 0.30 , 0.35 , 0.40 mev , respectively ; see also fig . 5 of ref .
the cranked hfb state is most suitable for describing the states with spin around its expectation value .
in fact , the difference of moments of inertia obtained with the two sets of frequencies is small in the spin range @xmath84 ; the results of configuration - mixing mainly differs in the lower spin region , @xmath96 .
it may be necessary to include the cranked hfb states with lower spin - expectation values for the configuration - mixing of the s - band in order to obtain the result that is independent of the detailed choice of the frequencies .
thus , the description of the s - band is not as simple as in the case of the g - band , especially in the low - spin region , @xmath96 .
it may not , however , be a serious problem because the s - bands for such low - spin parts have not been observed in experiments and it is difficult to compare with experimental data . in any case
, we need to study further for more satisfactory description of the s - band .
we have investigated the rotational bands in the rare earth nuclei by employing our recently developed microscopic framework , the angular - momentum - projected multi - cranked configuration - mixing method @xcite . in this method several cranked hfb states are utilized with a suitably chosen set of rotational frequencies .
we use the gogny force with the d1s parameter set as an effective interaction , and there is no ambiguity for the hamiltonian . we first apply our method to the g - band of various selected nuclei in the rare earth region .
reasonably good overall agreements are obtained for the energy spectra and the moments of inertia up to about @xmath97 . in a few cases
the moments of inertia at low spin are considerably overestimated and the increase of the inertia as a function of spin is not enough compared with the experimental data .
it is found that the selfconsistently calculated pairing correlations are too weak for such nuclei ; the average pairing gaps for both neutrons and protons are only about 70% of or even less than the even - odd mass differences .
if the pairing properties are nicely reproduced , the agreements of the moments of inertia are found to be excellent . in this way
we have confirmed that our method is capable to reliably describe the nuclear rotational motion near the ground state .
next we apply our approach to the study of the s - band in the nucleus @xmath0er for the first time .
the method of calculation is the same for the s - band ; the only difference is that the cranked hfb states with higher rotational frequencies are employed for the configuration - mixing , in which the two quasineutrons align their angular momenta .
thus the g- and s - bands can be calculated separately without the inter - band mixing between them .
the band crossing between the g- and s - bands can be reproduced , although the spin value , at which the two bands cross , is slightly larger than the observed one .
the calculated moment of inertia of the s - band is overestimated especially at high - spin states .
this is mainly because the selfconsistently calculated pairing correlation for neutrons vanishes in the cranked hfb states due to the alignment of two quasineutrons , which can not be avoided as long as the gogny d1s force is employed .
it is found that the result of configuration - mixing weakly depends on the choice of the set of cranking frequencies for the s - band , especially in the lower spin region , in contrast to the case of the g - band , where the result is independent of the choice of frequencies .
thus , further investigation is necessary for the proper description of the s - band , which is an important future work .
this work is supported in part by grant - in - aid for scientific research ( c ) no .
25@xmath98949 from japan society for the promotion of science . | recently we have proposed a reliable method to describe the rotational band in a fully microscopic manner .
the method has recourse to the configuration - mixing of several cranked mean - field wave functions after the angular - momentum - projection . by applying the method with the gogny d1s force as an effective interaction
, we investigate the moments of inertia of the ground state rotational bands in a number of selected nuclei in the rare earth region . as another application
we try to describe , for the first time , the two - neutron aligned band in @xmath0er , which crosses the ground state band and becomes the yrast states at higher spins .
fairly good overall agreements with the experimental data are achieved ; for nuclei , where the pairing correlations are properly described , the agreements are excellent .
this confirms that the previously proposed method is really useful for study of the nuclear rotational motion . |
a novel type of extensive air shower ( eas ) array ( the prisma project ) has been proposed some years ago @xcite to study cosmic ray spectrum and mass composition in the `` knee '' region .
existing experimental data in the knee region contradict each other and new approaches to the so - called `` knee problem '' are needed to solve this complicated and old problem .
the prisma project is based on an idea that the main eas component - hadrons has to be measured first of all .
special detectors ( en - detectors ) have been developed for this purpose .
it was proposed to combine the central part of the prisma array with the lhaaso project @xcite by introducing the en - detectors in the center of lhaaso array .
the en - detectors will make the lhaaso array sensitive to the hadronic eas component thus making it more powerful and informative .
we used the corsika program @xcite ( ver .
6.900 ) for eas simulations with qgsjet and gheisha models . calculations were performed for two primaries : proton and iron and for two altitudes : near sea level ( 170 m a.s.l . ) and high mountain 4300 m a.s.l . as a first step we made calculations for fixed primary energies from 10tev through 10pev and zenith angles @xmath0 .
we present here results for the number of hadrons and produced by them thermal neutrons inside a ring of 50 m radius around the eas axis as a function of primary energy .
the distributions over these numbers are also obtained . as any other monte - carlo program for eas simulation , corsika can not process particles with very low energies .
we used following cuts for particles : 50mev for hadrons , 0.5gev for muons , 60kev for electrons and gammas .
the mean number of evaporation neutrons @xmath1 produced by hadrons in 3-m layer of surrounding soil and/or construction materials has been calculated using an empirical relationship between them and the parent hadron energy ( in gev ) : @xmath2 this relationship originated from secondary particles production in hadronic interactions and was obtained taking into account experimental data @xcite and atomic mass a dependence of neutron production : @xmath3@xcite .
therefore , the total number of produced secondary neutrons should be summarized over all hadrons : @xmath4 it is easy to see that due to slow dependence on hadron energy and due to slow change of hadron mean energy with primary energy , @xmath5 can be simplified to : @xmath6 the latter means that the total number of evaporation neutrons produced in eas should be more or less proportional to the number of high - energy hadrons reaching the observation level .
the great bulk of these neutrons are thermalized later . thus recording thermal neutrons by detectors spread over big enough area one could recover the number of hadrons in eas
this idea is a basis of the prisma project .
mean numbers of produced neutrons and hadrons inside a ring of 50 m as a function of primary energy are shown in fig . 1 for primary proton and iron and for near sea level .
as one can see , all dependencies can be fitted with power law functions .
all indices for protons are close to 1 : @xmath7 at sea level and @xmath8 at 4300 m. for iron the indeces are little bit steeper at these energies .
small difference between primary proton and iron at the highest energy makes us sure that recovery of primary energy would be more adequate in the knee region .
one can also see that the mean number of neutrons at sea level and at 4300 m a.s.l .
differ by a factor of @xmath9 .
the latter means that at yangbajing level the array threshold energy can be lower by a factor of @xmath10 .
mean numbers of hadrons and secondary evaporation neutrons inside ring of 50 m radius as a function of primary energy at sea level for proton and iron primaries and at 4300 m for primary proton.,width=321 ] there are shown in fig .
2 the distributions of different eas components numbers inside 50 m around the axis for 2 primary energies : 30tev and 1pev .
an interesting issue here is rather high mean number of produced neutrons in comparison with hadrons and muons .
the latter means that thermal neutrons could be a better primary energy estimator than muons or high energy hadrons .
figure 2 also presents that at high altitude , within the primary energy range 30tev-10pev , the secondaries have higher quantities and then lead to higher energy resolution than ones at sea level .
it is necessary to test these simulation results at different altitudes .
we have made several tests with the prisma prototypes in moscow .
our test results are consistent with the simulations . a prototype of the prisma project array ( protoprisma ) @xcite has been developed and started running on a base of the nevod - decor detector at national research nuclear university mephi in moscow ( 170 m a.s.l .
it consists now of 32 inorganic scintillator @xmath11 situated inside the experimental building at a level of 4th floor just around the nevod water pool .
the neutron recording efficiency of the en - detector is equal to @xmath12 .
the detectors have a cylindrical shape with the scintillator area equal to 0.36 @xmath13 .
it is now our standard en - detector made on a base of commercial polyethylene ( pe ) water tank of 200 liters volume .
the scintillator thin sheets are situated on its bottom and are viewed by a single 6 pmt ( feu-200 ) .
the scintillator compound zns(ag)+ lif enriched with @xmath14 up to @xmath15 is very effective scintillator for heavy particle detection .
it produces 160000 photons per a neutron capture through the reaction @xmath14(n , a)t + 4.8 mev .
it allows us to collect more than 50 photoelectrons from pmt photo - cathode per n - capture . due to thin scintillator layer ( 30 @xmath16 )
, a single relativistic charged particle produces very small signal .
but , in a case of eas passage , correlated signals from many particles are summarized and can be measured .
therefore , the same detectors are used to measure two eas components : hadronic ( neutrons ) and electromagnetic .
that is why we call them as en - detectors ( electron - neutron ) .
the array consists now of 2 clusters of 16 detectors each .
the clusters have their own triggering system and work independently .
coincidence of 2 or more hit detectors within time of 2 @xmath17s produces a simple trigger and on - line program preprocesses data , selects eas events and marks them in accordance with different mathematical trigger conditions .
all pulses are integrated with a time constant of 1 @xmath17s and are digitized using 10 bits flash adc with a step of 1 @xmath17s . in a case of powerful enough eas , full pulse waveform of 20 ms duration is stored .
for energy deposit measurement in a wide range we use 2 signals from each detectors : from the last 12th dynode and from an intermediate 7th dynode as well . due to this
we have the energy range more than @xmath18 .
delayed pulses from thermal neutrons captures are measured in each detector in the time gate 0.1 - 20 ms . in order to test our simulations at sea level ,
we have processed the experimental data accumulated for 2 winter months with the aim to check eq.(1 ) normalization factor .
it is obvious that absolute value of neutron yield depends on the experimental conditions . in our case
the array is not situated in open air , but inside the building .
therefore , the result can be regarded as an estimation only .
detailed neutron yield measurements will be performed later when the next prisma prototype will be running in yangbajing area in real conditions .
mean number of recorded neutrons as a function of eas size ( @xmath19 ) is shown in fig . 3 in comparison with full - scale monte - carlo simulations made using corsika ( with the parameters mentioned above ) and the protoprisma array simulations .
for neutron production we use a formula similar to eq.(1 ) but with a normalization factor taking the our neutron recording efficiency into account .
the efficiency can be estimated as follows : @xmath20 , where @xmath21 is a probability for neutrons produced in soil ( concrete ) to escape to air @xcite and @xmath22=0.028 is a ratio of total detectors area ( 13.5 @xmath13 ) to the array area ( 484.5 @xmath13)(coverage ) . taking these efficiencies into account one can transform eq.(2 ) to : @xmath23 this relationship was used in the simulations of the experiment . the normalization factor ( c )
was varied to adjust it to the experimental points .
points in fig.3 were obtained with c=0.625 .
agreement is rather well ( note that due to limited statistics for the highest eas energies corresponding calculated points at the right tail are little bit lower ) .
mean number of recorded neutrons as a function of eas size for moscow site .
histogram - experiment , points - simulations.,width=321 ] as one can see , the mean number of recorded neutrons in the protoprisma experiment is @xmath24 1.6 times lower than expected one
. the reason could be found in the experimental details : detectors are placed inside building ( not in open air ) , the structure of the array is asymmetrical and the detector coverage in 2 clusters are different @xcite and finally , an existence of a large water pool inside the array affects undoubtedly on the neutrons yield making it lower .
in the frames of russian - chinese collaboration monte - carlo eas simulations have been performed and absolute normalization to the experiment has been done .
we obtained rather low yield of neutrons at the protoprisma location in moscow due to specifics of the experiment .
nevertheless , even at such yield and even at sea level it is possible to obtain some new results using the novel type of eas array .
we are looking forward to make such experiment at high altitude in yangbajing region .
it would make us possible to check the method at high altitude , to make calibration using the argo - ybj facilities and to measure real neutron yield in the future experiment site .
the present calculations show that the number of produced neutrons at yangbajing level is by a factor of 10 higher than that at sea level in eas of equal energy .
authors are grateful to russian foundation for basic research ( grants 09 - 02 - 12380-ofi - m and 11 - 02 - 01479a ) , to the neutrino program of russian academy of sciences , the federal target program scientific and educational cadres for innovative russia .
authors are grateful to the national natural science foundation of china ( nsfc ) ( the grant no . 10835007 . ) | the results of monte - carlo simulations of extensive air shower are presented to show the difference of hadronic component content at various altitudes with the aim to choose an optimal altitude for the prisma - like experiment .
corsika program for eas simulations with qgsjet and gheisha models was used to calculate the number of hadrons reaching the observational level inside a ring of 50 m radius around the eas axis . then the number of neutrons produced by the hadronic component was calculated using an empirical relationship between the two components .
we have tested the results with the protoprisma array at sea level , and recorded neutrons are close to the simulation results . _
keywords : eas , detector , neutrons , altitudes _ + |
a _ rack _ is a set @xmath4 with a binary operation @xmath5 satisfying that ( 1 ) for any @xmath6 and @xmath7 of @xmath4 , there is a unique element @xmath8 with @xmath9 , and ( 2 ) for all @xmath10 , @xmath11 .
a _ quandle _ is a rack satisfying that ( 3 ) @xmath12 for all @xmath13 .
( the condition ( 2 ) is called ( right ) self - distributivity .
refer to @xcite for details on algebras with this property . ) in @xcite , the _ knot quandle _
( or the _ fundamental quandle _ ) of a knot is defined and it is proved that the knot quandle is a complete invariant of a classical knot up to orientations of knots and the ambient space .
these notions are defined not only for classical knots but also for ( pl and locally flat , or smooth ) proper and oriented @xmath2-submanifolds of an oriented @xmath14-manifold ( cf .
quandles are very useful for studying oriented knots in @xmath0 and oriented surface knots in @xmath1 ( cf.@xcite ) . in @xcite
the notion of a presentation of a rack or a quandle is established and it is shown that a presentation of the knot quandle of an oriented knot is obtained from a diagram ( cf .
@xcite ) . in this paper
we show how to obtain a presentation of the knot quandle of a closed oriented @xmath2-submanifold of @xmath3 . in @xcite the notions of symmetric quandles , knot symmetric quandles ( or the fundamental symmetric quandles )
, colorings and quandle homological invariants using symmetric quandles for knots in @xmath0 and surface knots in @xmath1 were introduced and studied .
however presentations of symmetric quandles were not explicitly discussed .
the purpose of this paper is to establish the notion of presentations of symmetric quandles , as an analogy of that of presentations of quandles due to @xcite .
we define the notion of the _ knot symmetric quandle _ of a proper @xmath2-submanifold of an @xmath14-manifold , and show how to get a presentation of the knot symmetric quandle of a closed @xmath2-submanifold of @xmath3 from its diagram .
let @xmath4 be a quandle or a rack .
we follow the notation due to fenn and rourke @xcite . for example , @xmath15 means the unique element @xmath8 with @xmath16 .
an involution @xmath17 is a _ good involution _ if @xmath18 for every @xmath6 and @xmath7 of @xmath4 .
a _ symmetric quandle _ ( or a _ symmetric rack _ , resp . ) is a pair @xmath19 of a quandle ( or a rack , resp . )
@xmath4 and a good involution @xmath20 .
let @xmath4 be a trivial quandle , i.e. , @xmath21 for all @xmath22 .
then every involution of @xmath4 is a good involution .
conversely , a quandle such that every involution is a good involution is a trivial quandle ( cf .
@xcite ) . let @xmath4 be a kei ( @xcite ) , i.e , it is a quandle satisfying @xmath23 for all @xmath22 . then the identity map @xmath20 is a good involution .
conversely a quandle such that the identity map is a good involution is a kei ( cf .
@xcite ) . in general
, a kei has good involutions besides the identity map .
however , the dihedral quandle whose cardinarity is an odd integer has only the identity map as a good involution .
good involutions of dihedral quandles are classified in @xcite .
let @xmath24 be the conjugation quandle of a group @xmath25 , which is @xmath25 itself as a set and the quandle operation is the conjugation ; @xmath26 .
let @xmath20 be the invertion ; @xmath27 .
then @xmath19 is a symmetric quandle ( cf .
@xcite ) . [ doublecover ] for a quandle ( or a rack ) @xmath4 ,
let @xmath28 be a copy of @xmath4 and let @xmath29 be the disjoint union of @xmath4 and @xmath28 .
by @xmath30 , we mean the element of @xmath28 corresponding to an element @xmath6 of @xmath4 .
define a binary operations on @xmath29 as follows : @xmath31 where @xmath22 .
let @xmath32 be the involution interchanging @xmath6 and @xmath30 @xmath33 .
then @xmath34 is a symmetric quandle ( or a symmetric rack)(cf .
let @xmath35 be a ( pl and locally flat or smooth ) proper @xmath2-submanifold of an @xmath14-manifold @xmath36 .
let @xmath37 be a tubular neighborhood of @xmath35 and @xmath38 the exterior of @xmath35 , and take a base point @xmath39 of @xmath40 .
we call a pair @xmath41 a _ tadpole _ of @xmath35 if @xmath42 is an oriented meridian disk of @xmath35 and @xmath43 is an arc in the exterior @xmath40 starting from a point of @xmath44 and ending at the base point @xmath39 .
let @xmath45 \}$ ] be the quandle consisting of all homotopy classes of tadpoles with the operation @xmath46^{[(d_2 , a_2 ) ] } : = [ ( d_1 , a_1 \cdot a_2^{-1 } \cdot \partial d_2 \cdot a_2)]$ ] .
let @xmath47 be the map sending @xmath48 $ ] to @xmath49 $ ] , where @xmath50 means @xmath42 with the reverse orientation .
it is easily verified that @xmath51 is a good involution ( cf .
figure [ fig01.eps ] for @xmath52 ) , and we have a symmetric quandle @xmath53 . ] when @xmath36 is connected , the isomorphism class of @xmath54 does not depend on the base point @xmath39 .
so we denote it by @xmath55 . by an _ @xmath2-manifold knot _ in @xmath3 , we mean a closed ( pl and locally flat or smooth ) @xmath2-submanifold of @xmath3 .
let @xmath35 be a ( pl and locally flat or smooth ) proper @xmath2-submanifold of an @xmath14-manifold @xmath36 . the _ full knot quandle _ of @xmath35 is the quandle @xmath54 .
the _ knot symmetric quandle _ ( or the _ fundamental symmetric quandle _ ) of @xmath35 is the symmetric quandle @xmath53 , which we denote by @xmath56 .
when @xmath35 is an @xmath2-manifold knot in @xmath3 , the full knot quandle and the knot symmetric quandle of @xmath35 are denoted by @xmath57 and @xmath58 , respectively .
when @xmath35 is an orientable and oriented @xmath2-manifold knot in @xmath3 , if @xmath42 is an oriented meridian disk of @xmath35 whose orientation together with the orientation of @xmath35 matches the orientation of @xmath3 , we call a tadpole @xmath41 a _ positive tadpole _ of @xmath35 . the _ knot quandle _ or the _ positive knot quandle _ of @xmath35 is the quandle @xmath59 consisting of all homotopy classes of positive tadpoles ( cf .
@xcite ) .
let @xmath35 be an orientable and oriented @xmath2-manifold knot in @xmath3 .
the knot quandle @xmath59 is a subquandle of the full knot quandle @xmath57 , and the knot symmetric quandle @xmath60 is naturally identified with the double cover @xmath61 of the knot quandle @xmath59 in the sense of example [ doublecover ] .
the notion of the associated group @xmath62 of a quandle or a rack @xmath4 is generalized to the notion of the associated group of a symmetric quandle or a symmetric rack .
the _ associated group _
@xmath63 of a quandle or a rack @xmath4 is defined as @xmath64 [ thm : knotgrouppresentationa ] let @xmath35 be an oriented @xmath2-manifold knot in @xmath3 .
the associated group @xmath65 of the knot quandle @xmath59 is isomorphic to the knot group @xmath66 . to the knot group @xmath67
let @xmath35 be an oriented knot in @xmath0 and @xmath42 a diagram of @xmath35 in @xmath68 .
as shown in @xcite , associated with the diagram @xmath42 , we have a quandle @xmath69 defined by a presentation whose generators correspond to the arcs of the diagram and the relations correspond to the crossings .
for example , @xmath70 for the diagram in figure [ fig04.eps ] . the notion of a presentation of a quandle
is recalled in section [ sect : presentation ] .
[ thm : knotqdlepresentationa ] let @xmath35 be an oriented knot in @xmath0 and @xmath42 a diagram of @xmath35 in @xmath68
. then @xmath59 is isomorphic to @xmath69 .
replacement of @xmath71 with @xmath72 changes a presentation of a rack or a quandle @xmath4 to a group presentation of the associated group @xmath62 ( cf .
@xcite ) . for the previous example
, we have @xmath73 where @xmath71 stands for @xmath72 .
this is a wirtinger presentation of the knot group associated with the diagram .
theorem [ thm : knotqdlepresentationa ] is generalized to the higher dimensional case .
let @xmath35 be an @xmath2-manifold knot in @xmath3 .
we assume that it is in general position with respect to the projection @xmath74 .
the singular point set @xmath75 , that is the closure of the multiple point set @xmath76 in @xmath77 ( or its preimage in @xmath35 ) , is regarded as an @xmath78-dimensional stratified complex .
it is well known that the @xmath78-dimensional strata consist of transverse double points ( cf .
we call them the
_ double point strata _ and denote by @xmath79 the union of them .
the lower dimensional strata are in general complicated .
( when @xmath80 , the @xmath81-dimensional strata consists of triple points and branch points . ) for our purpose , classification of lower dimensional strata is not required at all since they do not contribute to the knot quandle . a _ diagram _ of @xmath35 is a subset of @xmath77 that is obtained from @xmath82 by removing a regular neighborhood of the singular set @xmath75 and put back the neighborhood of the upper branch of each double point stratum .
( see figure [ fig05.eps ] for local pictures of a diagram when @xmath80 . )
the connected components of a diagram called the _ sheets_. ( when @xmath83 , they are also called the _ arcs _ of a diagram . ) now we assume that @xmath35 is orientable and oriented . then all sheets of the diagram @xmath42 are also oriented .
we usually present the orientation of each sheet by a normal vector in @xmath77 as follows : let @xmath6 be a point of a sheet and let @xmath84 be an @xmath2-tuple of tangent vectors at @xmath6 representing the orientation of the sheet . take a normal vector @xmath85 such that @xmath86 matches the orientation of @xmath77 .
then we say that @xmath85 is the normal vector presenting the orientation of the sheet .
let @xmath35 be an oriented @xmath2-manifold knot in @xmath3 , and @xmath42 a diagram of @xmath35 in @xmath77 .
let @xmath69 be a quandle defined by a presentation whose generators correspond to the sheets of the diagram and the relations correspond to the double point strata . when the orientations of sheets are presented by normal vectors in @xmath77 as in the left of figure [ fig06.eps
] , the relation corresponding a double point stratum depicted in the right of figure [ fig06.eps ] is @xmath87 .
( this definition is compatible with the definitions in @xcite for the case where @xmath35 is an oriented knot in @xmath0 or an oriented surface knot in @xmath1 . )
[ thm : knotqdlepresentationb ] let @xmath35 be an oriented @xmath2-manifold knot in @xmath3 , and @xmath42 a diagram of @xmath35 in @xmath77
. then @xmath59 is isomorphic to @xmath69 .
this theorem is proved in section [ sect : proof ] . for a non - orientable @xmath2-manifold knot @xmath35 in @xmath3 ,
the knot quandle @xmath59 is not defined and we should consider the full knot quandle @xmath57 .
it will turn out that the knot symmetric quandle @xmath60 is more natural and practical than the full knot quandle @xmath57 itself .
for a symmetric quandle or a symmetric rack , the associated group is defined as follows .
let @xmath19 be a symmetric quandle or a symmetric rack .
the _ associated group _
@xmath88 of @xmath89 is defined by @xmath90 [ thm : knotqdlepresentationgroup ] for an @xmath2-manifold knot @xmath35 of @xmath3 , the associated group @xmath91 of the knot symmetric quandle @xmath92 is isomorphic to the knot group @xmath93 .
we will explain how to get a presentation of the knot symmetric quandle @xmath92 from a diagram later .
theorem [ thm : knotqdlepresentationgroup ] is proved in section [ sect : proof ] .
let @xmath35 be a standard 2-knot in @xmath1 .
the knot quandle and the full knot quandle are @xmath94 where @xmath95 $ ] and @xmath96 $ ] .
they are a trivial quandle consisting of a single element and a trivial quandle consisting of @xmath97 elements , respectively .
the associated groups of them are @xmath98 and @xmath99 . on the other hand ,
the knot symmetric quandle @xmath100 is @xmath92 where the involution @xmath51 is given by @xmath101 .
it has a presentation , as a symmetric quandle , @xmath102 and the associated group @xmath91 has a group presentation @xmath103 , which is @xmath98 .
-sphere ] as seen in this example and in theorem [ thm : knotgrouppresentationa ] , for an oriented @xmath2-manifold knot , the knot quandle @xmath59 is sufficient to recover the knot group .
however this is not the case for a non - orientable @xmath2-manifold knot .
let @xmath35 be a standard projective plane in @xmath1 .
since it is non - orientable , we can not define the knot quandle @xmath59 .
the full knot quandle is @xmath104 and the knot symmetric quandle @xmath100 is @xmath92 , where @xmath51 in the involution with @xmath105 .
the knot symmetric quandle @xmath100 has a presentation , as a symmetric quandle , @xmath106 and the associated group @xmath91 has a group presentation @xmath107 , which is @xmath108 .
now we explain how to get a presentation of the knot symmetric quandle from a diagram .
let @xmath35 be an @xmath2-manifold knot in @xmath3 , and @xmath42 a diagram of @xmath35 in @xmath77 .
we assume that @xmath35 is in general position with respect to the projection @xmath74 .
as stated before , a diagram is obtained from @xmath82 by removing a regular neighborhood of the singular set @xmath75 and by putting back the neighborhood of the upper branches of the double point strata .
semi - sheet _ of the diagram means a connected component of @xmath82 removed the neighborhood of @xmath75 ( without putting back the neighborhood of the upper branches of the double point strata ) .
each semi - sheet is a compact _ orientable _ @xmath2-manifold in @xmath77 , even if @xmath35 is non - orientable ( cf .
lemma 1 of @xcite ) . and semi - sheets ]
let @xmath109 be the semi - sheets of @xmath42 .
give them normal vectors @xmath110 arbitrarily .
we associate with each double point stratum an _
@xmath111-relation _ as in fig .
[ fig09.eps ] and a _
@xmath112-relation _ as in fig .
[ fig10.eps ] as follows : let @xmath113 and @xmath114 be the upper semi - sheets . if the normal vectors @xmath115 and @xmath116 of the upper semi - sheets @xmath113 and @xmath114 are coherent ( or incoherent , resp . ) , then let an @xmath111-relation be @xmath117 ( or @xmath118 , resp . ) .
let @xmath113 be an upper semi - sheet , and let @xmath119 and @xmath120 be the lower semi - sheets such that the normal vector of @xmath113 is directed from @xmath119 to @xmath120 .
if the normal vectors @xmath121 and @xmath122 of the lower semi - sheets @xmath119 and @xmath120 are coherent ( or incoherent , resp . ) , then let a @xmath112-relation be @xmath123 ( or @xmath124 , resp . ) . a _ symmetric quandle presentation associated with the diagram @xmath42 _ is a presentation whose generators are @xmath109 and the relations are @xmath111-relations and @xmath112-relations for double point strata .
we denote by @xmath125 the symmetric quandle with this presentation .
-relation ] -relation ] for a double point stratum where the normal vectors @xmath115 and @xmath116 of the upper semi - sheets are incoherent as depicted in the right of figure [ fig09.eps ] , there are two possibilities for an @xmath111-relation : one is @xmath118 and the other is @xmath126 . since @xmath51 is an involution , these two relations are equivalent
. similarly , there are two possibilities for a @xmath112-relation .
however they are equivalent up to the @xmath111-relation . in the right of figure [ fig10.eps
] , let @xmath114 be the over semi - sheet that is not labelled in the figure . for example
, consider the case depicted in the right of the figure and assume that the normal vector of @xmath114 is directed from @xmath120 to @xmath119 .
then we also have a relation @xmath127 as a @xmath112-relation .
the two relations @xmath128 and @xmath127 are equivalent up to the @xmath111-relation ; @xmath129 @xmath130 @xmath131 @xmath130 @xmath132 @xmath130 @xmath133 @xmath130 @xmath134
. thus it is sufficient for defining the symmetric quandle @xmath125 to consider a single @xmath111-relation and a single @xmath112-relation for each double point stratum .
let @xmath42 be the diagram of a trefoil knot depicted in figure [ fig11.eps ] .
( the dotted arcs mean the neighborhood of the upper branches of the double point strata . )
choose normal vectors for the semi - sheets as in the figure .
we may take @xmath135 , @xmath136 and @xmath137 as @xmath111-relations , and @xmath138 , @xmath139 and @xmath140 as @xmath112-relations .
the symmetric quandle @xmath125 associated with the diagram has a presentation @xmath141 by eliminating @xmath142 and @xmath143 from this presentation , we have @xmath144 the relation @xmath136 is equivalent to @xmath145 , and the relation @xmath146 becomes @xmath147 , which is equivalent to @xmath148 .
the relation @xmath149 becomes @xmath150 , which is equivalent to @xmath151 and hence to @xmath152 .
thus we have a presentation @xmath153 for @xmath125 .
[ thm : knotqdlepresentationc ] let @xmath35 be an @xmath2-manifold knot in @xmath3 , and @xmath42 a diagram of @xmath35 in @xmath77 . the knot symmetric quandle @xmath154 of @xmath35 is isomorphic to the symmetric quandle @xmath125 associated with the diagram @xmath42 .
we will prove this theorem in section [ sect : proof ] . as a corollary
, we see that the knot symmetric quandle @xmath154 of an @xmath2-manifold knot in @xmath3 has a finite presentation .
let @xmath35 be an @xmath2-manifold knot in @xmath3 .
let @xmath19 be a symmetric quandle .
an _ @xmath19-coloring _ of a diagram of @xmath35 is an assignment of an element of @xmath4 to each semi - sheet of the diagram satisfying an a - relation and a b - relation of each double point stratum .
we denote by @xmath155 the set of @xmath19-colorings of @xmath42 , and by @xmath156 the set of homomorphisms from the knot symmetric quandle @xmath154 to the symmetric quandle @xmath19 .
there is a bijection from @xmath155 to @xmath156 . in particular ,
the cardinarity @xmath157 is a knot invariant of @xmath35 . using a symmetric quandle , some invariants derived from a quandle , as quandle colorings , quandle cocycle invariants @xcite , etc
can be also applied for non - orientable surface knots in @xmath1 ( cf .
some other related researches are found in @xcite .
this section is devoted to introducing the notion of a presentation of a symmetric quandle .
first we recall the notions of a presentation of a rack and a presentation of a quandle due to fenn and rourke @xcite .
let @xmath158 be a non - empty set and let @xmath159 be the free group generated by the elements of @xmath158 .
for an element @xmath160 , we often denote @xmath161 by @xmath162 .
the _ free rack _ is a rack @xmath163 with a rack operation @xmath164 .
an element @xmath165 of @xmath166 is also denoted by @xmath167 ; @xmath168 .
moreover we assume that the free group @xmath159 acts on @xmath169 from the right by @xmath170 .
then @xmath166 is regarded as an augmented rack with an augmentation @xmath171 given by @xmath172 .
( see page 355 of @xcite for the definition of an augmented rack . )
let @xmath173 be a subset of @xmath174 .
let @xmath175 be the smallest congruence containing @xmath173 with respect to axioms of a rack . start with the set @xmath173 and enlarge @xmath173 by repeating the following moves : * add every diagonal element @xmath176 . * if @xmath177 then add @xmath178 . * if @xmath177 and @xmath179 then add @xmath180 .
* if @xmath177 then add @xmath181 and @xmath182 for every @xmath183 . * if @xmath177 then add @xmath184 and @xmath185 for every @xmath186 .
the result is @xmath175 , which we call the _ set of rack consequences _ of @xmath173 .
let @xmath187 this is the rack whose generating set is @xmath158 and the defining relation set is @xmath173 . for a rack @xmath4 ,
if there is a map @xmath188 which induces an isomorphism @xmath189 then we call it a _ rack presentation_. for a quandle presentation , let @xmath190 be the smallest congruence containing @xmath173 with respect to axioms of a quandle
. start with the given set @xmath191 and enlarge it by repeating the moves ( e1 ) , ( e2 ) , ( e3 ) , ( r1 ) , ( r2 ) above and the following ( q ) .
* add @xmath192 for every @xmath183 .
the result is @xmath190 , which we call the _ set of quandle consequences _ of @xmath173 .
let @xmath193 this is the quandle whose generating set is @xmath158 and the defining relation set is @xmath173 . for a quandle @xmath4 ,
if there is a map @xmath188 which induces an isomorphism @xmath194 then we call it a _ quandle presentation_. note that @xmath195 , and @xmath196 ( cf
. remark of p. 365 of @xcite ) .
the _ free quandle _ on @xmath158 is @xmath197 .
now we define presentations of a symmetric rack and a symmetric quandle .
first we define the free symmetric rack .
let @xmath158 be a non - empty set , and let @xmath198 be a copy of @xmath158 in which @xmath199 stands for the copy of an element @xmath183 .
( we assume @xmath200 . )
let @xmath201 be the free group generated by @xmath202 and @xmath203 the normal closure of @xmath204 in the free group .
the quotient @xmath205 is naturally identified with the free group @xmath159 by @xmath206 and @xmath207 . for an element @xmath160 , we often denote @xmath161 by @xmath162 . the _ free symmetric rack _ on @xmath158 is a symmetric rack whose underlying rack is @xmath208 with a rack operation @xmath209 for @xmath210 and @xmath211 , and the involution @xmath51 is given by @xmath212 .
we denote the free symmetric rack @xmath213 by @xmath214 for short .
an element @xmath215 is also denoted by @xmath167 .
the image @xmath216 is denoted by @xmath217 or @xmath218 , which is also by @xmath219 or @xmath220 .
the free group @xmath159 acts on @xmath214 by @xmath221 .
then @xmath214 is regarded as an augmented rack with an augmentation @xmath222 given by @xmath172 .
we define an _ augmented symmetric rack _ to be a symmetric rack @xmath223 whose underlying rack @xmath4 is an augmented rack with an augmentation @xmath224 such that the involution @xmath225 is compatible with the canonical involution @xmath226 of the group @xmath25 .
the free symmetric rack @xmath227 with the action of @xmath159 is an augmented symmetric rack .
let @xmath173 be a subset of @xmath228 .
let @xmath229 be the smallest congruence containing @xmath173 with respect to axioms of a symmetric rack
. start with the set @xmath173 and enlarge @xmath173 by repeating the following moves : * add every diagonal element @xmath176 . * if @xmath177 then add @xmath178 . * if @xmath177 and @xmath179 then add @xmath180 . * if @xmath177 then add @xmath181 and @xmath182 for every @xmath183 . * if @xmath177 then add @xmath184 and @xmath185 for every @xmath230 . * if @xmath177 then add @xmath231 .
the result is @xmath229 , which we call the _ set of symmetric rack consequences _ of @xmath173 .
let @xmath232 this is the symmetric rack whose generating set is @xmath158 and the defining relation set is @xmath173 . for a symmetric rack @xmath233 ,
if there is a map @xmath188 which induces an isomorphism @xmath234 then we call it a _ symmetric rack presentation_. for a symmetric quandle presentation ,
let @xmath235 be the smallest congruence containing @xmath173 with respect to axioms of a symmetric quandle
. start with the given set @xmath236 and enlarge it by repeating the moves ( se1 ) , ( se2 ) , ( se3 ) , ( sr1 ) , ( sr2 ) , ( s ) above and the move ( q ) : * add @xmath192 for every @xmath183 .
the result is @xmath235 , which we call the _ set of symmetric quandle consequences _ of @xmath173 .
let @xmath237 this is the symmetric quandle whose generating set is @xmath158 and the defining relation set is @xmath173 .
note that @xmath238 and that @xmath239 .
the _ free symmetric quandle _ on @xmath158 is @xmath240 .
we denote it by @xmath241 .
in this section we prove theorems @xmath242 , @xmath243 and @xmath244 .
_ proof of theorem [ thm : knotqdlepresentationc]_. we use a similar argument as in the proof of theorem 2 of @xcite .
let @xmath245 be the semi - sheets of the diagram @xmath42 , and let @xmath246 be normal vectors assigned for the semi - sheets .
let @xmath247 be a regular neighborhood of the singular set @xmath75 and put @xmath248 , which is the union of the semi - sheets . for each @xmath251
, consider an oriented small interval @xmath252 in @xmath77 intersecting the interior of the semi - sheet @xmath113 in the same direction with @xmath115 , and let @xmath253 \subset { { \mathbb r}}^{n+1 } \times { { \mathbb r}}= { { \mathbb r}}^{n+2}$ ] , which is a meridian disk of @xmath35 .
we give an orientation to @xmath254 such that @xmath255 is oriented in the same direction with @xmath252 . by this orientation , we assume that @xmath256 is an oriented meridian disk of @xmath35 .
let @xmath257 be a straight line in @xmath3 starting from a point of @xmath258 to the base point @xmath39 .
we denote by the same symbol @xmath113 the element of @xmath57 represented by the pair @xmath259 , and also by the same symbol @xmath113 the element of the knot group @xmath260 represented by the loop @xmath261 .
let @xmath262 , and let @xmath263 the free symmetric quandle on @xmath4 and @xmath264 the free group on @xmath4 .
we can obtain a symmetric quandle homomorphism @xmath265 with @xmath266 $ ] and @xmath267 $ ] for @xmath251 , and a group homomorphism @xmath268 with @xmath269 $ ] and @xmath270 $ ] for @xmath251 . put @xmath271 and @xmath272 .
then @xmath273 is a proper submanifold of @xmath274 . by the argument of lemma 4
( 2 ) of @xcite , we see that the knot group @xmath275 is identified with the free group @xmath264 on @xmath4 and by using also the argument in @xcite , we see that the knot symmetric quandle @xmath276 is identified with the free symmetric quandle @xmath263 on @xmath4 .
we continue the proof . put @xmath277 , where @xmath79 is the union of the double point strata and @xmath278 is the union of the lower dimensional strata .
we divide the regular neighborhood @xmath247 as follows : let @xmath279 be a regular neighborhood of @xmath278 in @xmath77 , and put @xmath280 .
let @xmath281 be a regular neighborhood of @xmath282 , and put @xmath283 .
we assume @xmath284 .
let @xmath285 and @xmath286 be arcs in a cylinder @xmath287 $ ] as in figure [ fig12.eps ] .
note that @xmath288 is a trivial @xmath289-bundle over @xmath282 , since each component of @xmath79 has a trivialization determined from the four sheets around it .
identify @xmath288 with @xmath290 and @xmath291 $ ] with @xmath292)$ ] .
we may assume that @xmath35 restricted to @xmath291 $ ] is @xmath293 in @xmath292)$ ] .
then we see that for each double point stratum , we have an @xmath111-relation and a @xmath112-relation .
therefore , the homomorphisms @xmath294 and @xmath295 induces a symmetric quandle homomorphism @xmath296 with @xmath266 $ ] and @xmath267 $ ] for @xmath251 , and a group homomorphism @xmath297 with @xmath269 $ ] and @xmath270 $ ] for @xmath251 . in @xcite
it is proved by using the van kampen theorem that the group homomorphism @xmath298 is an isomorphism from the fact that @xmath278 has codimension more than two in @xmath77 .
a similar argument is applied for quandles ( cf .
@xcite ) and we see that the quandle homomorphism @xmath299 is an isomorphism .
in the proof of theorem [ thm : knotqdlepresentationc ] , we used the van kampen theorem for quandles in @xcite to obtain the fact that @xmath299 is an isomorphism .
this fact can be also proved by a similar argument with the proof of theorem 4.7 in @xcite .
_ proof of theorem [ thm : knotqdlepresentationgroup]_. by theorem [ thm : knotqdlepresentationc ] , the knot symmetric quandle @xmath58 has a presentation associated with a diagram @xmath42 . in the proof of theorem
[ thm : knotqdlepresentationc ] ( or by theorem 2 of @xcite ) , we can obtain a group presentation of the knot group @xmath300 associated with the diagram @xmath42 , that is obtained from the symmetric quandle presentation associated with @xmath42 by replacing the quandle operation @xmath71 with conjugation @xmath72 , and @xmath301 with @xmath302 .
this is the associated group of the symmetric quandle @xmath58 .
_ proof of theorem [ thm : knotqdlepresentationb]_. in the proof of theorem [ thm : knotqdlepresentationc ] , take the normal vectors @xmath246 such that they present the orientation of the semi - sheets .
then all @xmath111-relations are in the case of @xmath117 as in the left of figure [ fig09.eps ] .
thus we can reduce the generating set from the set of semi - sheets to the set of sheets of the diagram .
the @xmath112-relations are as in the left of figure [ fig10.eps ] .
then we see that the knot quandle @xmath59 has a quandle presentation associated with the oriented diagram @xmath42 .
s. kamada , _ quandles with good involutions , their homologies and knot invariants _ , in : intelligence of low dimensional topology 2006 , pp . 101108 , ser .
knots everything , vol .
40 , world sci .
hackensack , nj , 2007 . | a symmetric quandle is a quandle with a good involution . for a knot in @xmath0 , a knotted surface in @xmath1 or an @xmath2-manifold knot in @xmath3 ,
the knot symmetric quandle is defined .
we introduce the notion of a symmetric quandle presentation , and show how to get a presentation of a knot symmetric quandle from a diagram . [ section ] [ theorem]definition [ theorem]corollary [ theorem]remark [ theorem]example [ theorem]lemma [ theorem]proposition |
a simple cosmological model , a flat universe with nearly scale - invariant adiabatic gaussian fluctuations , has proven to be a remarkably good fit to ever improving cosmic microwave background ( cmb ) data @xcite , large - scale structure data @xcite , supernova data @xcite , cluster measurements @xcite , distance measurements @xcite , measurements of strong @xcite and weak @xcite gravitational lensing effects .
observations of cmb have been playing an essential role in testing the model and constraining its basic parameters .
the _ wmap _
satellite @xcite has been measuring temperature and polarization anisotropies of the cmb over the full sky since 2001 . with 7 years of integration ,
the errors in the temperature spectrum at each multipole are dominated by cosmic variance ( rather than by noise ) up to @xmath12 , and the signal - to - noise at each multipole exceeds unity up to @xmath13 @xcite .
the power spectrum of primary cmb on smaller angular scales has been measured by other experiments up to @xmath14 @xcite .
the polarization data show the most dramatic improvements over our earlier _ wmap _ results : the temperature - polarization cross power spectra measured by _ wmap _ at @xmath15 are still dominated by noise , and the errors in the 7-year cross power spectra have improved by nearly 40% compared to the 5-year cross power spectra . while the error in the power spectrum of the cosmological @xmath8-mode polarization @xcite averaged over @xmath16 is cosmic - variance limited
, individual multipoles are not yet cosmic - variance limited .
moreover , the cosmological @xmath10-mode polarization has not been detected @xcite . the temperature - polarization ( te and tb ) power spectra offer unique tests of the standard model .
the te spectrum can be predicted given the cosmological constraints from the temperature power spectrum , and the tb spectrum is predicted to vanish in a parity - conserving universe .
they also provide a clear physical picture of how the cmb polarization is created from quadrupole temperature anisotropy .
we show the success of the standard model in an even more striking way by measuring this correlation in map space , rather than in harmonic space .
the constraints on the basic 6 parameters of a flat @xmath1cdm model ( see table [ tab : summary ] ) , as well as those on the parameters beyond the minimal set ( see table [ tab : deviation ] ) , continue to improve with the 7-year _ wmap _ temperature and polarization data , combined with improved external astrophysical data sets . in this paper
, we shall give an update on the cosmological parameters , as determined from the latest cosmological data set .
our best estimates of the cosmological parameters are presented in the last columns of table [ tab : summary ] and [ tab : deviation ] under the name `` _ wmap_+bao+@xmath0 . '' while this is the minimal combination of robust data sets such that adding other data sets does not significantly improve most parameters , the other data combinations provide better limits than _ wmap_+bao+@xmath0 in some cases .
for example , adding the small - scale cmb data improves the limit on the primordial helium abundance , @xmath17 ( see table [ tab : yhe ] and section [ sec : helium ] ) , the supernova data are needed to improve limits on properties of dark energy ( see table [ tab : darkenergy ] and section [ sec : darkenergy ] ) , and the power spectrum of luminous red galaxies ( lrgs ; see section [ sec : lrg ] ) improves limits on properties of neutrinos ( see footnotes g , h , and i in table [ tab : deviation ] and sections [ sec : massnu ] and [ sec : neff ] ) .
the cmb can also be used to probe the abundance as well as the physics of clusters of galaxies , via the sz effect @xcite . in this paper
, we shall present the _ wmap _
measurement of the averaged profile of sz effect measured towards the directions of known clusters of galaxies , and discuss implications of the _ wmap _ measurement for the very small - scale ( @xmath18 ) power spectrum recently measured by the south pole telescope ( spt ; * ? ? ?
* ) and atacama cosmology telescope ( act ; * ? ? ?
* ) collaborations .
this paper is one of six papers on the analysis of the _ wmap _ 7-year data : @xcite report on the data processing , map - making , and systematic error limits ; @xcite on the modeling , understanding , and subtraction of the temperature and polarized foreground emission ; @xcite on the measurements of the temperature and polarization power spectra , extensive testing of the parameter estimation methodology by monte carlo simulations , and the cosmological parameters inferred from the _ wmap _ data alone ; @xcite on the assessments of statistical significance of various `` anomalies '' in the _ wmap _ temperature map reported in the literature ; and @xcite on _ wmap _ s measurements of the brightnesses of planets and various celestial calibrators .
this paper is organized as follows . in section [ sec : pol ] , we present results from the new method of analyzing the polarization patterns around temperature hot and cold spots . in section [ sec
: analysis ] , we briefly summarize new aspects of our analysis of the _ wmap _
7-year temperature and polarization data , as well as improvements from the 5-year data . in section [ sec : parameters ] , we present updates on various cosmological parameters , except for dark energy .
we explore the nature of dark energy in section [ sec : darkenergy ] . in section [ sec : ng ] , we present limits on primordial non - gaussianity parameters @xmath19 . in section [ sec : sz
] , we report detection , characterization , and interpretation of the sz effect toward locations of known clusters of galaxies .
we conclude in section [ sec : conclusion ] .
lccccc primary & @xmath20 & 2.227 & 2.253 & @xmath21 & @xmath22 & @xmath23 & 0.1116 & 0.1122 & @xmath24 & @xmath25 & @xmath26 & 0.729 & 0.728 & @xmath27 & @xmath28 & @xmath29 & 0.966 & 0.967 & @xmath30 & @xmath31 & @xmath32 & 0.085 & 0.085 & @xmath33 & @xmath34 & @xmath35 & @xmath36 & @xmath36 & @xmath37 & @xmath38 derived & @xmath39 & 0.809 & 0.810 & @xmath40 & @xmath41 & @xmath0 & @xmath42 & @xmath43 & @xmath44 km / s / mpc & @xmath45 km / s / mpc & @xmath46 & 0.0451 & 0.0455 & @xmath47 & @xmath48 & @xmath49 & 0.226 & 0.226 & @xmath50 & @xmath51 & @xmath52 & 0.1338 & 0.1347 & @xmath53 & @xmath54 & @xmath55 & 10.4 & 10.3 & @xmath56 & @xmath56 & @xmath57 & 13.79 gyr & 13.76 gyr & @xmath58 gyr & @xmath59 gyr [ tab : summary ] lccccc section [ sec : gw ] & grav .
wave & no running ind . &
@xmath60 & @xmath61 & @xmath62 section [ sec : running ] & running index & no grav . wave & @xmath63 & @xmath64 & @xmath65 section [ sec : ok ] & curvature & @xmath66 & n / a & @xmath67 & @xmath68 section [ sec : ad ] & adiabaticity & axion & @xmath69 & @xmath70 & @xmath71 & & curvaton & @xmath72 & @xmath73 & @xmath74 section [ sec : tb ] & parity violation & chern - simons & @xmath75 & n / a & n / a section [ sec : massnu ] & neutrino mass & @xmath66 & @xmath76 & @xmath77 & @xmath78 & & @xmath79 & @xmath80 & @xmath81 & @xmath76 section [ sec : neff ] & relativistic species & @xmath66 & @xmath82 & n / a & @xmath83 section [ sec : ng ] & gaussianity & local & @xmath84 & n / a & n / a & & equilateral & @xmath85 & n / a & n / a & & orthogonal & @xmath86 & n / a & n / a [ tab : deviation ] lcc @xmath17 & @xmath87 ( 95% cl ) & @xmath6 ( 68% cl ) [ tab : yhe ] lccccc section [ sec : w_flat ] & @xmath88 & constant @xmath89 & @xmath90 & @xmath91 & @xmath92 section [ sec : w_curve ] & @xmath93 & constant @xmath89 & @xmath94 & @xmath95 & @xmath96 & & @xmath97 & @xmath98 & @xmath99 & @xmath100 & & & + @xmath0+sn & + bao+@xmath0+sn & + bao+@xmath0+@xmath101+sn section [ sec : w0wa ] & @xmath88 & @xmath102 & @xmath103 & @xmath104 & @xmath105 & & @xmath106 & @xmath107 & @xmath108 & @xmath109 [ tab : darkenergy ]
electron - photon scattering converts quadrupole temperature anisotropy in the cmb at the decoupling epoch , @xmath7 , into linear polarization @xcite .
this produces a correlation between the temperature pattern and the polarization pattern @xcite .
different mechanisms for generating fluctuations produce distinctive correlated patterns in temperature and polarization : * adiabatic scalar fluctuations predict a radial polarization pattern around temperature cold spots and a tangential pattern around temperature hot spots on angular scales greater than the horizon size at the decoupling epoch , @xmath110 . on angular scales smaller than the sound horizon size at the decoupling epoch , _ both _ radial _ and _ tangential patterns are formed around both hot and cold spots , as the acoustic oscillation of the cmb modulates the polarization pattern @xcite . as we have not seen any evidence for non - adiabatic fluctuations (
* see section [ sec : ad ] for the 7-year limits ) , in this section we shall assume that fluctuations are purely adiabatic .
* tensor fluctuations predict the opposite pattern : the temperature cold spots are surrounded by a tangential polarization pattern , while the hot spots are surrounded by a radial pattern @xcite .
since there is no acoustic oscillation for tensor modes , there is no modulation of polarization patterns around temperature spots on small angular scales .
we do not expect this contribution to be visible in the _ wmap _ data , given the noise level .
* defect models predict that there should be minimal correlations between temperature and polarization on @xmath111 @xcite .
the detection of large - scale temperature polarization fluctuations rules out any causal models as the primary mechanism for generating the cmb fluctuations @xcite .
this implies that the fluctuations were either generated during an accelerating phase in the early universe or were present at the time of the initial singularity .
this section presents the first direct measurement of the predicted pattern of adiabatic scalar fluctuations in cmb polarization maps .
we stack maps of stokes @xmath112 and @xmath113 around temperature hot and cold spots to show the expected polarization pattern at the statistical significance level of 8@xmath9 .
while we have detected the te correlations in the first year data @xcite , we present here the direct real space pattern around hot and cold spots . in section [ sec : polarizationdiscussion ] , we discuss the relationship between the two measurements .
we first identify temperature hot ( or cold ) spots , and then stack the polarization data ( i.e. , stokes @xmath112 and @xmath113 ) on the locations of the spots .
as we shall show below , the resulting polarization data is equivalent to the temperature _ peak_-polarization correlation function which is similar to , but different in an important way from , the temperature - polarization correlation function . our definitions of stokes @xmath112 and @xmath113 follow that of @xcite : the polarization that is parallel to the galactic meridian is @xmath116 and @xmath117 .
starting from this , the polarization that is rotated by @xmath118 from east to west ( clockwise , as seen by an observer on earth looking up at the sky ) has @xmath119 and @xmath120 , that perpendicular to the galactic meridian has @xmath121 and @xmath117 , and that rotated further by @xmath118 from east to west has @xmath119 and @xmath122 . with one more rotation
we go back to @xmath116 and @xmath117 .
we show this in figure [ fig : directions ] . and @xmath113 .
we use galactic coordinates with north up and east left . in this example
, @xmath114 is always negative , and @xmath115 is always zero .
when @xmath123 and @xmath124 , the polarization pattern is radial .
, width=264 ] as the predicted polarization pattern around temperature spots is either radial or tangential , we find it most convenient to work with @xmath114 and @xmath115 first introduced by @xcite : @xmath125 these transformed stokes parameters are defined with respect to the new coordinate system that is rotated by @xmath126 , and thus they are defined with respect to the line connecting the temperature spot at the center of the coordinate and the polarization at an angular distance @xmath127 from the center ( also see figure [ fig : directions ] ) .
note that we have used the small - angle ( flat - sky ) approximation for simplicity of the algebra .
this approximation is justified as we are interested in relatively small angular scales , @xmath128 .
the above definition of @xmath114 is equivalent to the so - called `` tangential shear '' statistic used by the weak gravitational lensing community . by following what has been already done for the tangential shear
, we can find the necessary formulae for @xmath114 and @xmath115 .
specifically , we shall follow the derivations given in @xcite . with the small - angle approximation , @xmath112 and @xmath113
are related to the @xmath8- and @xmath10-mode polarization in fourier space @xcite as @xmath129e^{i{\mathbf l}\cdot{\bm{\theta}}},\\ -u(\bm{\theta } ) & = & \int\frac{d^2{\mathbf l}}{(2\pi)^2 } \left[e_{\mathbf l}\sin(2\varphi)+b_{\mathbf l}\cos(2\varphi)\right]e^{i{\mathbf l}\cdot{\bm{\theta}}},\end{aligned}\ ] ] where @xmath130 is the angle between @xmath131 and the line of galactic latitude , @xmath132 .
note that we have included the negative signs on the left hand side because our sign convention for the stokes parameters is opposite of that used in equation ( 38 ) of @xcite .
the transformed stokes parameters are given by @xmath133\right.\\ & & \qquad\qquad \left.+b_{\mathbf l}\sin[2(\phi-\varphi)]\right\}e^{i{\mathbf l}\cdot{\bm{\theta}}},\\ \nonumber -u_r(\bm{\theta } ) & = & \int\frac{d^2{\mathbf l}}{(2\pi)^2 } \left\{e_{\mathbf l}\sin[2(\phi-\varphi)]\right.\\ & & \qquad\qquad \left .- b_{\mathbf l}\cos[2(\phi-\varphi)]\right\}e^{i{\mathbf l}\cdot{\bm{\theta}}}.\end{aligned}\ ] ] the stacking of @xmath114 and @xmath115 at the locations of temperature _ peaks _ can be written as @xmath134 where the angle bracket , @xmath135 , denotes the average over the locations of peaks , @xmath136 is the surface number density of peaks ( of the temperature fluctuation ) at the location @xmath137 , @xmath138 is the total number of temperature peaks used in the stacking analysis , and @xmath139 is equal to 0 at the masked pixels and 1 otherwise . defining the density contrast of peaks , @xmath140 , we find @xmath141 where @xmath142 is the fraction of sky outside of the mask , and we have used @xmath143 . in appendix [ app :
stacking ] , we use the statistics of peaks of gaussian random fields to relate @xmath144 to the temperature-@xmath8-mode polarization cross power spectrum @xmath145 , @xmath146 to the temperature-@xmath10-mode polarization cross power spectrum @xmath147 , and the stacked temperature profile , @xmath148 , to the temperature power spectrum @xmath149 .
we find @xmath150 where @xmath151 and @xmath152 are the harmonic transform of window functions , which are a combination of the experimental beam , pixel window , and any other additional smoothing applied to the temperature and polarization data , respectively , and @xmath153 is the `` scale - dependent bias '' of peaks found by @xcite averaged over peaks . see appendix [ app : stacking ] for details .
what do @xmath154 and @xmath155 look like ?
the @xmath114 map is expected to be non - zero for a cosmological signal , while the @xmath115 map is expected to vanish in a parity - conserving universe unless some systematic error rotates the polarization plane uniformly . to understand the shape of @xmath114 as well as its physical implications ,
let us begin by showing the smoothed @xmath145 spectra and the corresponding temperature-@xmath114 correlation functions , @xmath156 , in figure [ fig : clte ] .
( note that @xmath157 and @xmath158 can be computed from equations ( [ eq : qr ] ) and ( [ eq : ur ] ) , respectively , with @xmath159 and @xmath160 . )
this shows three distinct effects causing polarization of cmb ( see * ? ? ?
* for a pedagogical review ) : * @xmath161 , where @xmath162 is the angular size of the radius of the horizon size at the decoupling epoch . using the comoving horizon size of @xmath163 gpc and the comoving angular diameter distance to the decoupling epoch of @xmath164 gpc as derived from the _ wmap _ data , we find @xmath165 . as
this scale is so much greater than the sound horizon size ( see below ) , only gravity affects the physics .
suppose that there is a newtonian gravitational potential , @xmath166 , at the center of a perturbation , @xmath167 .
if it is overdense at the center , @xmath168 , and thus it is a cold spot according to the sachs
wolfe formula @xcite , @xmath169 .
the photon fluid in this region will flow into the gravitational potential well , creating a converging flow .
such a flow creates the quadrupole temperature anisotropy around an electron at @xmath170 , producing polarization that is radial , i.e. , @xmath123 .
since the temperature is negative , we obtain @xmath171 , i.e. , anti - correlation @xcite . on the other hand ,
if it is overdense at the center , then the photon fluid moves outward , producing polarization that is tangential , i.e. , @xmath172 .
since the temperature is positive , we obtain @xmath171 , i.e. , anti - correlation .
the anti - correlation at @xmath170 is a smoking - gun for the presence of super - horizon fluctuations at the decoupling epoch @xcite , which has been confirmed by the _ wmap
_ data @xcite .
* @xmath173 , where @xmath174 is the angular size of the radius of the _ sound _ horizon size at the decoupling epoch . using the comoving sound horizon size of @xmath175 gpc and @xmath164 gpc as derived from the _ wmap _ data , we find @xmath176 .
again , consider a potential well with @xmath168 at the center . as the photon fluid flows into the well
, it compresses , increasing the temperature of the photons . whether or not this increase can reverse the sign of the temperature fluctuation ( from negative to positive ) depends on whether the initial perturbation was adiabatic .
if it was adiabatic , then the temperature would reverse sign at @xmath177 .
note that the photon fluid is still flowing in , and thus the polarization direction is radial , @xmath123 .
however , now that the temperature is positive , the correlation reverses sign : @xmath178 . a similar argument ( with the opposite sign )
can be used to show the same result , @xmath179 , for @xmath180 at the center . as an aside
, the temperature reverses sign on smaller angular scales for isocurvature fluctuations .
* @xmath181 .
again , consider a potential well with @xmath168 at the center . at @xmath182 ,
the pressure of the photon fluid is so great that it can slow down the flow of the fluid .
eventually , at @xmath183 , the pressure becomes large enough to reverse the direction of the flow ( i.e. , the photon fluid expands ) . as a result
the polarization direction becomes tangential , @xmath172 ; however , as the temperature is still positive , the correlation reverses sign again : @xmath184 . on even smaller scales ,
the correlation reverses sign again ( see figure 2 of * ? ? ?
* ) because the temperature gets too cold due to expansion .
we do not see this effect in figure [ fig : clte ] because of the smoothing .
lastly , there is no correlation between @xmath185 and @xmath114 at @xmath167 because of symmetry .
these features are essentially preserved in the peak - polarization correlation as measured by the stacked polarization profiles .
we show them in figure [ fig : cpk ] for various values of the threshold peak heights .
the important difference is that , thanks to the scale - dependent bias @xmath186 , the small - scale trough at @xmath187 is enhanced , making it easier to observe . on the other hand ,
the large - scale anti - correlation is suppressed .
we can therefore conclude that , with the _
wmap_data , we should be able to measure the compression phase at @xmath188 , as well as the reversal phase at @xmath189 .
we also show the profiles calculated from numerical simulations ( gray solid lines ) .
the agreement with equation ( [ eq : qr ] ) is excellent .
we also show the predicted profiles of the stacked temperature data in figure [ fig : cpt ] .
we use the foreground - reduced v@xmath190w temperature map at the healpix resolution of @xmath191 to find temperature peaks .
first , we smooth the foreground - reduced temperature maps in 6 differencing assemblies ( das ) ( v1 , v2 , w1 , w2 , w3 , w4 ) to a common resolution of @xmath192 ( fwhm ) using @xmath193 where @xmath194 is the appropriate beam transfer function for each da @xcite , and @xmath195 $ ] is the pixel window function for @xmath196 , @xmath197 , times the spherical harmonic transform of a gaussian with @xmath198 . we then coadd the foreground - reduced v- and w - band maps with the inverse noise variance weighting , and remove the monopole from the region outside of the mask ( which is already negligibly small , @xmath199k ) . for the mask ,
we combine the new 7-year _ kq85 _ mask , _ kq85y7 _ ( defined in * ? ? ?
* also see section [ sec : analysis_wmap ] ) and _ p06 _ masks , leaving 68.7% of the sky available for the analysis .
we find the locations of minima and maxima using the software `` hotspot '' in the healpix package @xcite . over the full sky ( without the mask )
, we find 20953 maxima and 20974 minima . as the maxima and minima found by hotspot still contain negative and positive peaks , respectively
, we further select the `` hot spots '' by removing all negative peaks from maxima , and the `` cold spots '' by removing all positive peaks from minima .
this procedure corresponds to setting the threshold peak height to @xmath200 ; thus , our prediction for @xmath201 is the top - left panel of figure [ fig : cpk ] . outside of the mask , we find 12387 hot spots and 12628 cold spots .
the r.m.s .
temperature fluctuation is @xmath202 .
what does the theory predict ?
using equation ( [ eq : npk ] ) with the power spectrum @xmath203 $ ] where @xmath204 is the noise bias of the v@xmath190w map before gaussian smoothing and @xmath205 is the 5-year best - fitting power - law @xmath1cdm temperature power spectrum , we find @xmath206 for @xmath200 and @xmath207 ; thus , the number of observed hot and cold spots is consistent with the predicted number .
if we ignore the noise bias ; thus , even with a gaussian smoothing , the contribution from noise is not negligible . ] as for the polarization data , we use the _ raw _
( i.e. , without foreground cleaning ) polarization maps in v and w bands .
we have checked that the cleaned maps give similar results with slightly larger error bars , which is consistent with the excess noise introduced by the template foreground cleaning procedure @xcite .
as we are focused on relatively small angular scales , @xmath208 , in this analysis , the results presented in this section would not be affected by a potential systematic effect causing an excess power in the w - band polarization data on large angular scales , @xmath209 .
however , note that this excess power could just be a statistical fluctuation @xcite .
we form two sets of the data : ( i ) v , w , and v@xmath190w band maps smoothed to a common resolution of @xmath192 , and ( ii ) v , w , and v@xmath190w band maps without any additional smoothing .
the first set is used only for visualization , whereas the second set is used for the @xmath210 analysis .
we extract a square region of @xmath211 around each temperature hot or cold spot .
we then coadd the extracted @xmath185 images with uniform weighting , and @xmath112 and @xmath113 images with the inverse noise variance weighting .
we have eliminated the pixels masked by _
kq85y7 _ and _ p06 _ from each @xmath212 region when we coadd images , and thus the resulting stacked image has the smallest noise at the center ( because the masked pixels usually appear near the edge of each image ) .
we also accumulate the inverse noise variance per pixel as we coadd @xmath112 and @xmath113 maps .
the coadded inverse noise variance maps of @xmath112 and @xmath113 will be used to estimate the errors of the stacked images of @xmath112 and @xmath113 per pixel , which will then be used for the @xmath210 analysis .
we find that the stacked images of @xmath112 and @xmath113 have constant offsets , which is not surprising .
since these affect our determination of polarization directions , we remove monopoles from the stacked images of @xmath112 and @xmath113 .
the size of each pixel in the stacked image is @xmath213 , and the number of pixels is @xmath214 . finally , we compute @xmath114 and @xmath115 from the stacked images of stokes @xmath112 and @xmath113 using equations ( [ eq : qrdef ] ) and ( [ eq : urdef ] ) , respectively .
lccc hot , @xmath112 , v@xmath190w & 661.9 & @xmath215 & @xmath216 hot , @xmath113 , v@xmath190w & 661.1 & @xmath217 & @xmath218 hot , @xmath114 , v@xmath190w & 694.2 & @xmath219 & @xmath220 hot , @xmath115 , v@xmath190w & 629.2 & @xmath221 & @xmath222 cold , @xmath112 , v@xmath190w & 668.3 & @xmath223 & @xmath224 cold , @xmath113 , v@xmath190w & 682.7 & @xmath225 & @xmath226 cold , @xmath114 , v@xmath190w & 682.2 & @xmath227 & @xmath228 cold , @xmath115 , v@xmath190w & 657.8 & @xmath229 & @xmath230 hot , @xmath112 , v@xmath231w & 559.8 & & hot , @xmath113 , v@xmath231w & 629.8 & & hot , @xmath114 , v@xmath231w & 662.2 & & hot , @xmath115 , v@xmath231w & 567.0 & & cold , @xmath112 , v@xmath231w & 584.0 & & cold , @xmath113 , v@xmath231w & 668.2 & & cold , @xmath114 , v@xmath231w & 616.0 & & cold , @xmath115 , v@xmath231w & 636.9 & & [ tab : chi2 ] in figure [ fig : minima ] and [ fig : maxima ] , we show the stacked images of @xmath185 , @xmath112 , @xmath113 , @xmath114 , and @xmath115 around temperature cold spots and hot spots , respectively .
the peak values of the stacked temperature profiles agree with the predictions ( see the dashed line in the top - left panel of figure [ fig : cpt ] ) .
a dip in temperature ( for hot spots ; a bump for cold spots ) at @xmath232 is clearly visible in the data .
while the stokes @xmath112 and @xmath113 measured from the data exhibit the expected features , they are still fairly noisy .
the most striking images are the stacked @xmath114 ( and @xmath185 ) .
the predicted features are clearly visible , particularly the compression phase at @xmath233 and the reversal phase at @xmath234 in @xmath114 : the polarization directions around temperature cold spots are radial at @xmath235 and tangential at @xmath236 , and those around temperature hot spots show the opposite patterns , as predicted .
how significant are these features ? before performing the quantitative @xmath210 analysis , we first compare @xmath114 and @xmath115 using both the ( v@xmath190w)/2 sum map ( here , v@xmath190w refers to the inverse noise variance weighted average ) as well as the ( v@xmath231w)/2 difference map ( bottom panels of figure [ fig : minima ] and [ fig : maxima ] ) . the @xmath114 map ( which is expected to be non - zero for a cosmological signal ) shows clear differences between the sum and difference maps , while the @xmath115 map ( which is expected to vanish in a parity - conserving universe unless some systematic error rotates the polarization plane uniformly ) is consistent with zero in both the sum and difference maps .
next , we perform the standard @xmath210 analysis .
we summarize the results in table [ tab : chi2 ] .
we report the values of @xmath210 measured with respect to zero signal in the second column , where the number of degrees of freedom ( dof ) is 625 . for each sum map combination
, we fit the data to the predicted signal to find the best - fitting amplitude .
the largest improvement in @xmath210 is observed for @xmath114 , as expected from the visual inspection of figure [ fig : minima ] and [ fig : maxima ] : we find @xmath219 and @xmath227 for the stacking of @xmath114 around hot and cold spots , respectively .
the improvement in @xmath210 is @xmath237 and @xmath228 , respectively ; thus , we detect the expected polarization patterns around hot and cold spots at the level of 5.4@xmath9 and 6@xmath9 , respectively .
the combined significance exceeds 8@xmath9 .
on the other hand , we do not find any evidence for @xmath115 .
the @xmath210 values with respect to zero signal per dof are @xmath238 ( hot spots ) and @xmath239 ( cold spots ) , and the probabilities of finding larger values of @xmath210 are 44.5% and 18% , respectively .
but , can we learn anything about cosmology from this result ?
while the standard model predicts @xmath240 and hence @xmath241 , models in which the global parity symmetry is violated can create @xmath242 @xcite .
therefore , we fit the measured @xmath115 to the predicted @xmath114 , finding a null result : @xmath243 and @xmath229 ( 68% cl ) , or equivalently @xmath244 and @xmath245 ( 68% cl ) for hot and cold spots , respectively . averaging these numbers ,
we obtain @xmath246 ( 68% cl ) , which is consistent with ( although not as stringent as ) the limit we find from the full analysis presented in section [ sec : tb ] . finally , all the @xmath210 values measured from the difference maps are consistent with a null signal .
how do these results compare to the full analysis of the te power spectrum ? by fitting the 7-year @xmath145 data to the same power spectrum used above ( 5-year best - fitting power - law @xmath1cdm model from @xmath247 to @xmath248 , i.e. , dof@xmath249777 ) , we find the best - fitting amplitude of @xmath250 and @xmath251 , i.e. , a 21@xmath9 detection of the te signal .
this is reasonable , as we used only the v- and w - band data for the stacking analysis , while we used also the q - band data for measuring the te power spectrum ; @xmath201 is insensitive to information on @xmath252 ( see top left panel of figure [ fig : cpk ] ) ; and the smoothing suppresses the power at @xmath253 ( see left panel of figure [ fig : clte ] ) .
nevertheless , there is probably a way to extract more information from @xmath201 by , for example , combining data at different threshold peak heights and smoothing scales .
if the temperature fluctuations of the cmb obey gaussian statistics and global parity symmetry is respected on cosmological scales , the temperature-@xmath8-mode polarization cross power spectrum , @xmath145 , contains all the information about the temperature - polarization correlation . in this sense
, the stacked polarization images do not add any new information .
the detection and measurement of the temperature-@xmath8 mode polarization cross - correlation power spectrum , @xmath145 @xcite , can be regarded as equivalent to finding the predicted polarization patterns around hot and cold spots . while we have shown that one can write the stacked polarization profile around temperature spots in terms of an integral of @xmath145 , the formal equivalence between this new method and @xmath145 is valid only when temperature fluctuations obey gaussian statistics , as the stacked @xmath112 and @xmath113 maps measure correlations between temperature _ peaks _ and polarization .
so far there is no convincing evidence for non - gaussianity in the temperature fluctuations observed by _
* see section [ sec : ng ] for the 7-year limits on primordial non - gaussianity , and bennett et al .
2010 for discussion on other non - gaussian features ) .
nevertheless , they provide striking confirmation of our understanding of the physics at the decoupling epoch in the form of radial and tangential polarization patterns at two characteristic angular scales that are important for the physics of acoustic oscillation : the compression phase at @xmath254 and the reversal phase at @xmath255 .
also , this analysis does not require any analysis in harmonic space , nor decomposition to @xmath8 and @xmath10 modes .
the analysis is so straightforward and intuitive that the method presented here would also be useful for null tests and systematic error checks
. the stacked image of @xmath115 should be particularly useful for systematic error checks .
any experiments that measure both temperature and polarization should be able to produce the stacked images such as presented in figure [ fig : minima ] and [ fig : maxima ] .
lccc high @xmath256 & te & detected at 21@xmath9 & detected at 13@xmath9 & tb & @xmath257 & @xmath258 low @xmath256 & ee & @xmath259 & @xmath260 & bb & @xmath261 ( 95% cl ) & @xmath262 ( 95% cl ) & ee / bb & @xmath263 ( 95% cl ) & @xmath264 ( 95% cl ) & tb / eb & @xmath265 & @xmath266 all @xmath256 & te / ee / bb & @xmath267 ( 95% cl ) & @xmath263 ( 95% cl ) & tb / eb & @xmath268 & @xmath269 [ tab : improvements ] * foreground mask*. the 7-year temperature analysis masks , _
kq85y7 _ and _ kq75y7 _ , have been slightly enlarged to mask the regions that have excess foreground emission , particularly in the hii regions gum and ophiuchus , identified in the difference between foreground - reduced maps at different frequencies . as a result ,
kq85y7 _ and _ kq75y7 _ masks eliminate an additional 3.4% and 1.0% of the sky , leaving 78.27% and 70.61% of the sky for the cosmological analyses , respectively .
see section 2.1 of @xcite for details . there is no change in the polarization _
p06 _ mask ( see section 4.2 of * ? ? ?
* for definition of this mask ) , which leaves 73.28% of the sky .
* point sources and the sz effect*. we continue to marginalize over a contribution from unresolved point sources , assuming that the antenna temperature of point sources declines with frequency as @xmath270 ( see equation ( 5 ) of * ? ? ?
the 5-year estimate of the power spectrum from unresolved point sources in q band in units of antenna temperature , @xmath271 , was @xmath272 @xcite , and we used this value and the error bar to marginalize over the power spectrum of residual point sources in the 7-year parameter estimation .
the subsequent analysis showed that the 7-year estimate of the power spectrum is @xmath273 @xcite , which is somewhat lower than the 5-year value because more sources are resolved by _ wmap _ and included in the source mask .
the difference in the diffuse mask ( between _ kq85y5 _ and _ kq85y7 _ ) does not affect the value of @xmath271 very much : we find @xmath274 instead of @xmath275 if we use the 5-year diffuse mask and the 7-year source mask .
the source power spectrum is sub - dominant in the total power .
we have checked that the parameter results are insensitive to the difference between the 5-year and 7-year residual source estimates .
we continue to marginalize over a contribution from the sz effect using the same template as for the 3- and 5-year analyses @xcite .
we assume a uniform prior on the amplitude of this template as @xmath276 , which is now justified by the latest limits from the spt collaboration , @xmath277 ( 68% cl ; * ? ? ?
* ) , and the act collaboration , @xmath278 ( 95% cl ; * ? ? ?
* high-@xmath256 temperature and polarization*. we increase the multipole range of the power spectra used for the cosmological parameter estimation from @xmath279 to @xmath280 for the tt power spectrum , and from @xmath281 to @xmath282 for the te power spectrum .
we use the 7-year v- and w - band maps @xcite to measure the high-@xmath256 tt power spectrum in @xmath283 .
while we used only q- and v - band maps to measure the high-@xmath256 te and tb power spectra for the 5-year analysis @xcite , we also include w - band maps in the 7-year high-@xmath256 polarization analysis . with these data ,
we now detect the high-@xmath256 te power spectrum at 21@xmath9 , compared to 13@xmath9 for the 5-year high-@xmath256 te data .
this is a consequence of adding two more years of data and the w - band data .
the tb data can be used to probe a rotation angle of the polarization plane , @xmath284 , due to potential parity - violating effects or systematic effects . with the 7-year high-@xmath256 tb data we find a limit @xmath257 ( 68% cl ) . for comparison ,
the limit from the 5-year high-@xmath256 tb power spectrum was @xmath285 ( 68% cl ; * ? ? ? * ) .
see section [ sec : tb ] for the 7-year limit on @xmath284 from the full analysis .
* low-@xmath256 temperature and polarization*. except for using the 7-year maps and the new temperature _ kq85y7 _ mask , there is no change in the analysis of the low-@xmath256 temperature and polarization data : we use the internal linear combination ( ilc ) map @xcite to measure the low-@xmath256 tt power spectrum in @xmath286 , and calculate the likelihood using the gibbs sampling and blackwell - rao ( br ) estimator @xcite . for the implementation of the br estimator in the 5-year analysis ,
see section 2.1 of @xcite .
we use ka- , q- , and v - band maps for the low-@xmath256 polarization analysis in @xmath287 , and evaluate the likelihood directly in pixel space as described in appendix d of @xcite . to get a feel for improvements in the low-@xmath256 polarization data with two additional years of integration
, we note that the 7-year limits on the optical depth , and the tensor - to - scalar ratio and rotation angle from the low-@xmath256 polarization data _ alone _ , are @xmath259 ( 68% cl ; see * ? ? ?
* ) , @xmath263 ( 95% cl ; see section [ sec : gw ] ) , and @xmath288 ( 68% cl ; see section [ sec : tb ] ) , respectively .
the corresponding 5-year limits were @xmath260 @xcite , @xmath264 ( see section [ sec : gw ] ) , and @xmath266 @xcite , respectively . in table
[ tab : improvements ] , we summarize the improvements from the 5-year data mentioned above . the _ wmap _ data are statistically powerful enough to constrain 6 parameters of a flat @xmath1cdm model with a tilted spectrum .
however , to constrain deviations from this minimal model , other cmb data probing smaller angular scales and astrophysical data probing the expansion rates , distances , and growth of structure are useful .
the best limits on the primordial helium abundance , @xmath17 , are obtained when the _ wmap _ data are combined with the power spectrum data from other cmb experiments probing smaller angular scales , @xmath289 .
we use the temperature power spectra from the arcminute cosmology bolometer array receiver ( acbar ; * ? ? ?
* ) and quest at dasi ( quad ) @xcite experiments .
for the former , we use the temperature power spectrum binned in 16 band powers in the multipole range @xmath290 . for the latter ,
we use the temperature power spectrum binned in 13 band powers in @xmath290 .
we marginalize over the beam and calibration errors of each experiment : for acbar , the beam error is 2.6% on a 5 arcmin ( fwhm ) gaussian beam and the calibration error is 2.05% in temperature . for quad ,
the beam error combines a 2.5% error on 5.2 and 3.8 arcmin ( fwhm ) gaussian beams at 100 ghz and 150 ghz , respectively , with an additional term accounting for the sidelobe uncertainty ( see appendix a of * ? ? ? * for details ) .
the calibration error is 3.4% in temperature .
the acbar data are calibrated to the _ wmap _
5-year temperature data , and the quad data are calibrated to the boomerang data @xcite which are , in turn , calibrated to the _ wmap_1-year temperature data .
( the quad team takes into account the change in the calibration from the 1-year to the 5-year _ wmap _ data . )
the calibration errors quoted above are much greater than the calibration uncertainty of the _ wmap _ 5-year data ( 0.2% * ? ? ?
this is due to the noise of the acbar , quad , and boomerang data . in other words , the above calibration errors are dominated by the statistical errors that are uncorrelated with the _ wmap _ data .
we thus treat the _ wmap _ , acbar , and quad data as independent .
figure [ fig : cl ] shows the _ wmap _
7-year temperature power spectrum @xcite as well as the temperature power spectra from acbar and quad .
we do not use the other , previous small - scale cmb data , as their statistical errors are much larger than those of acbar and quad , and thus adding them would not improve the constraints on the cosmological parameters significantly .
the new power spectrum data from the spt @xcite and act @xcite collaborations were not yet available at the time of our analysis .
there are two main astrophysical priors that we shall use in this paper : the hubble constant and the angular diameter distances out to @xmath291 and @xmath292 . * a gaussian prior on the present - day hubble constant , @xmath293 ( 68% cl ; * ? ? ?
the quoted error includes both statistical and systematic errors .
this measurement of @xmath0 is obtained from the magnitude - redshift relation of 240 low-@xmath294 type ia supernovae at @xmath295 .
the absolute magnitudes of supernovae are calibrated using new observations from _ hst _ of 240 cepheid variables in six local type ia supernovae host galaxies and the maser galaxy ngc 4258 .
the systematic error is minimized by calibrating supernova luminosities directly using the geometric maser distance measurements .
this is a significant improvement over the prior that we adopted for the 5-year analysis , @xmath296 , which is from the hubble key project final results @xcite .
* gaussian priors on the distance ratios , @xmath297 and @xmath298 , measured from the _ two - degree field galaxy redshift survey _
( _ 2dfgrs _ ) and the _ sloan digital sky survey _ data release 7 ( _ sdss _
dr7 ) @xcite .
the inverse covariance matrix is given by equation ( 5 ) of @xcite .
these priors are improvements from those we adopted for the 5-year analysis , @xmath299 and @xmath300 @xcite . + the above measurements can be translated into a measurement of @xmath301 at a single , `` pivot '' redshift : @xmath302 @xcite .
@xcite used the two - point correlation function of _
sdss_-dr7 lrgs to measure @xmath301 at @xmath303 .
they found @xmath304 , which is an excellent agreement with the above measurement by @xcite at a similar redshift .
the excellent agreement between these two independent studies , which are based on very different methods , indicates that the systematic error in the derived values of @xmath301 may be much smaller than the statistical error . + here
, @xmath305 is the comoving sound horizon size at the baryon drag epoch @xmath306 , @xmath307 for @xmath306 , we use the fitting formula proposed by @xcite .
the effective distance measure , @xmath308 @xcite , is given by @xmath309^{1/3},\ ] ] where @xmath310 is the proper ( not comoving ) angular diameter distance : @xmath311}{(1+z)\sqrt{\left|\omega _ { k}\right| } } , \label{eq : da}\ ] ] where @xmath312=\sin x$ ] , @xmath313 , and @xmath314 for @xmath315 ( @xmath316 ; positively curved ) , @xmath317 ( @xmath318 ; flat ) , and @xmath319 ( @xmath320 ; negatively curved ) , respectively .
the hubble expansion rate , which has contributions from baryons , cold dark matter , photons , massless and massive neutrinos , curvature , and dark energy , is given by equation ( [ eq : hubble ] ) in section [ sec : exactnu ] .
the cosmological parameters determined by combining the _ wmap _ data , bao , and @xmath0 will be called `` _ wmap_+bao+@xmath0 , '' and they constitute our best estimates of the cosmological parameters , unless noted otherwise . note that , when redshift is much less than unity , the effective distance approaches @xmath321 .
therefore , the effect of different cosmological models on @xmath308 do not appear until one goes to higher redshifts .
if redshift is very low , @xmath308 is simply measuring the hubble constant .
a combination of the _ wmap _ data and the power spectrum of luminous red galaxies ( lrgs ) measured from the _ sdss _ dr7 is a powerful probe of the total mass of neutrinos , @xmath322 , and the effective number of neutrino species , @xmath323 @xcite .
we thus combine the lrg power spectrum @xcite with the _ wmap _ 7-year data and the hubble constant @xcite to update the constraints on @xmath322 and @xmath323 reported in @xcite .
note that bao and the lrg power spectrum can not be treated as independent data sets because a part of the measurement of bao used lrgs as well .
the luminosity distances out to high-@xmath294 type ia supernovae have been the most powerful data for first discovering the existence of dark energy @xcite and then constraining the properties of dark energy , such as the equation of state parameter , @xmath89 ( see * ? ? ?
* for a recent review ) . with more than 400 type ia supernovae discovered ,
the constraints on the properties of dark energy inferred from type ia supernovae are now limited by systematic errors rather than by statistical errors .
there is an indication that the constraints on dark energy parameters are different when different methods are used to fit the light curves of type ia supernovae @xcite .
we also found that the parameters of the minimal 6-parameter @xmath1cdm model derived from two compilations of @xcite are different : one compilation uses the light curve fitter called salt - ii @xcite while the other uses the light curve fitter called mlcs2k2 @xcite .
for example , @xmath26 derived from _ wmap_+bao+salt - ii and _ wmap_+bao+mlcs2k2 are different by nearly @xmath324 , despite being derived from the same data sets ( but processed with two different light curve fitters ) .
if we allow the dark energy equation of state parameter , @xmath89 , to vary , we find that @xmath89 derived from _ wmap_+bao+salt - ii and _ wmap_+bao+mlcs2k2 are different by @xmath325 . at the moment it is not obvious how to estimate systematic errors and
properly incorporate them in the likelihood analysis , in order to reconcile different methods and data sets . in this paper , we shall use one compilation of the supernova data called the `` constitution '' samples @xcite .
the reason for this choice over the others , such as the compilation by @xcite that includes the latest data from the sdss - ii supernova survey , is that the constitution samples are an extension of the `` union '' samples @xcite that we used for the 5-year analysis ( see section 2.3 of * ? ? ?
. more specifically , the constitution samples are the union samples plus the latest samples of nearby type ia supernovae optical photometry from the center for astrophysics ( cfa ) supernova group ( cfa3 sample ; * ? ? ?
therefore , the parameter constraints from a combination of the _ wmap _ 7-year data , the latest bao data described above @xcite , and the constitution supernova data may be directly compared to the `` _
wmap_+bao+sn '' parameters given in table 1 and 2 of @xcite .
this is a useful comparison , as it shows how much the limits on parameters have improved by adding two more years of data .
however , given the scatter of results among different compilations of the supernova data , we have decided to choose the `` _
wmap_+bao+@xmath0 '' ( see section [ sec : wmap+bao+h0 ] ) as our best data combination to constrain the cosmological parameters , except for dark energy parameters . for dark energy parameters ,
we compare the results from _ wmap_+bao+@xmath0 and _ wmap_+bao+sn in section [ sec : darkenergy ] .
note that we always marginalize over the absolute magnitudes of type ia supernovae with a uniform prior .
can we measure angular diameter distances out to higher redshifts ?
measurements of gravitational lensing time delays offer a way to determine absolute distance scales @xcite . when a foreground galaxy lenses a background variable source ( e.g. , quasars ) and produces multiple images of the source , changes of the source luminosity due to variability appear on multiple images at different times .
the time delay at a given image position @xmath326 for a given source position @xmath327 , @xmath328 , depends on the angular diameter distances as ( see , e.g. , * ? ? ?
* for a review ) @xmath329 where @xmath330 , @xmath331 , and @xmath332 are the angular diameter distances out to a lens galaxy , to a source galaxy , and between them , respectively , and @xmath333 is the so - called fermat potential , which depends on the path length of light rays and gravitational potential of the lens galaxy . the biggest challenge for this method is to control systematic errors in our knowledge of @xmath333 , which requires a detailed modeling of mass distribution of the lens .
one can , in principle , minimize this systematic error by finding a lens system where the mass distribution of lens is relatively simple .
the lens system b1608 + 656 is not a simple system , with two lens galaxies and dust extinction ; however , it has one of the most precise time - delay measurements of quadruple lenses .
the lens redshift of this system is relatively large , @xmath334 @xcite .
the source redshift is @xmath335 @xcite .
this system has been used to determine @xmath0 to 10% accuracy @xcite .
@xcite have obtained more data from the deep _ hst _ advanced camera for surveys ( acs ) observations of the asymmetric and spatially extended lensed images , and constrained the slope of mass distribution of the lens galaxies .
they also obtained ancillary data ( for stellar dynamics and lens environment studies ) to control the systematics , particularly the the so - called `` mass - sheet degeneracy , '' which the strong lensing data alone can not break . by doing so
, they were able to reduce the error in @xmath0 ( including the systematic error ) by a factor of two @xcite .
they find a constraint on the `` time - delay distance , '' @xmath101 , as @xmath336 where the number is found from a gaussian fit to the likelihood of @xmath101 ; however , the actual shape of the likelihood is slightly non - gaussian .
we thus use : * likelihood of @xmath101 out to the lens system b1608 + 656 given by @xcite , @xmath337 } { \sqrt{2\pi}(x-\lambda)\sigma},\ ] ] where @xmath338 , @xmath339 , @xmath340 , and @xmath341 .
this likelihood includes systematic errors due to the mass - sheet degeneracy , which dominates the total error budget ( see section 6 of * ? ? ?
* for more details ) .
note that this is the only lens system for which @xmath101 ( rather than @xmath0 ) has been constrained .
, is the angular diameter distance to the lens , @xmath330 , multiplied by the distance ratio , @xmath342 , the sensitivity of @xmath101 to cosmological parameters is somewhat limited compared to that of @xmath330 @xcite . on the other hand ,
if the density profile of the lens galaxy is approximately given by @xmath343 , the observed einstein radius and velocity dispersion of the lens galaxy can be used to infer the same distance ratio , @xmath342 , and thus one can use this property to constrain cosmological parameters as well @xcite , up to uncertainties in the density profile @xcite . by combining measurements of the time - delay , einstein ring , and velocity dispersion , one can in principle measure @xmath330 directly , thereby turning strong gravitational lens systems into standard rulers @xcite . while the accuracy of the current data for b1608 + 656 does not permit us to determine @xmath330 precisely yet ( s. h. suyu and p. j. marshall 2009 , private communication ) , there seems to be exciting future prospects for this method .
future prospects of the time - delay method are also discussed in @xcite . ] when we evaluate the likelihood of external astrophysical data sets , we often need to compute the hubble expansion rate , @xmath344 . while we treated the effect of massive neutrinos on @xmath344 approximately for the 5-year analysis of the external data sets presented in @xcite
, we treat it exactly for the 7-year analysis , as described below .
the total energy density of massive neutrino species , @xmath345 , is given by ( in natural units ) @xmath346 where @xmath347 is the mass of each neutrino species .
using the comoving momentum , @xmath348 , and the present - day neutrino temperature , @xmath349 k , we write @xmath350 throughout this paper , we shall assume that all massive neutrino species have the equal mass @xmath351 , i.e. , @xmath352 for all @xmath353 .
galaxy redshift surveys or weak lensing surveys @xcite . ]
when neutrinos are relativistic , one may relate @xmath345 to the photon energy density , @xmath354 , as @xmath355 where @xmath323 is the effective number of neutrino species .
note that @xmath356 for the standard neutrino species . @xcite . ]
this motivates our writing equation ( [ eq : rhonu ] ) as @xmath357 where @xmath358 the limits of this function are @xmath359 for @xmath360 , and @xmath361 for @xmath362 , where @xmath363 is the riemann zeta function .
we find that @xmath364 can be approximated by the following fitting formula : is normalized by the density in the non - relativistic limit . here
, @xmath345 is normalized by the density in the relativistic limit .
both results agree with the same precision . ]
@xmath365^{1/p},\ ] ] where @xmath366 and @xmath367 .
this fitting formula is constructed such that it reproduces the asymptotic limits in @xmath360 and @xmath368 exactly .
this fitting formula underestimates @xmath364 by 0.1% at @xmath369 , and overestimates by 0.35% at @xmath370 .
the errors are smaller than these values at other @xmath371 s . using this result ,
we write the hubble expansion rate as @xmath372\right.\\ & & \left.+ \frac{\omega_k}{a^2 } + \frac{\omega_\lambda}{a^{3(1+w_{\rm eff}(a ) ) } } \right\}^{1/2 } , \label{eq : hubble}\end{aligned}\ ] ] where @xmath373 for @xmath374 k. using the massive neutrino density parameter , @xmath375 , for the standard 3 neutrino species , we find @xmath376 one can check that @xmath377 for @xmath378 .
one may compare equation ( [ eq : hubble ] ) , which is exact ( if we compute @xmath364 exactly ) , to equation ( 7 ) of @xcite , which is approximate . throughout this paper ,
we shall use @xmath26 to denote the dark energy density parameter at present : @xmath379 .
the function @xmath380 in equation ( [ eq : hubble ] ) is the effective equation of state of dark energy given by @xmath381 , and @xmath382 is the usual dark energy equation of state , i.e. , the dark energy pressure divided by the dark energy density : @xmath383 . for vacuum energy ( cosmological constant ) , @xmath89 does not depend on time , and @xmath66 .
lccccc section [ sec : gw ] & power - law & @xmath29 & @xmath30 & @xmath384 & @xmath31 section [ sec : running ] & running & @xmath29 & @xmath385 & @xmath386 & @xmath387 & & @xmath388 & @xmath389 & @xmath390 & @xmath391 section [ sec : gw ] & tensor & @xmath29 & @xmath392 & @xmath393 & @xmath394 & & @xmath395 & @xmath396 & @xmath397 & @xmath398 section [ sec : running ] & running & @xmath29 & @xmath399 & & @xmath400 & + tensor & @xmath395 & @xmath401 & n / a & @xmath401 & & @xmath388 & @xmath402 & & @xmath403 [ tab : ns ] , from the polarization data ( bb , ee and te ) alone .
all the other cosmological parameters , including the optical depth , are fixed at the 5-year best - fit @xmath1cdm model @xcite .
the vertical axis shows @xmath404 , where @xmath405 is the likelihood and @xmath406 is the maximum value .
this quantity may be interpreted as the standard @xmath210 , as the likelihood is approximately a gaussian near the maximum ; thus , @xmath407 corresponds to the 95.4% cl limit .
the solid , dashed and dot - dashed lines show the likelihood as a function of @xmath395 from the bb - only , bb@xmath190ee , and bb@xmath190ee@xmath190te data .
( left ) the 7-year polarization data .
we find @xmath261 , 1.6 , and 0.93 ( 95.4% cl ) from the bb - only , bb@xmath190ee , and bb@xmath190ee@xmath190te data , respectively .
( right ) the 5-year polarization data .
we find @xmath262 , 2.7 , and 1.6 ( 95.4% cl ) from the bb - only , bb@xmath190ee , and bb@xmath190ee@xmath190te data , respectively .
, width=302 ] the 7-year _ wmap _ data combined with bao and @xmath0 exclude the scale - invariant spectrum by 99.5% cl , if we ignore tensor modes ( gravitational waves ) . for a power - law power spectrum of primordial curvature perturbations
@xmath408 , i.e. , @xmath409 where @xmath410 , we find @xmath411 for comparison , the _ wmap _ data - only limit is @xmath412 @xcite , and the _ wmap _ plus the small - scale cmb experiments acbar @xcite and quad @xcite is @xmath413 . as explained in section 3.1.2 of @xcite , the small - scale cmb data do not reduce the error bar in @xmath29 very much because of relatively large statistical errors , beam errors , and calibration errors . how about tensor modes ? while the b - mode polarization is a smoking - gun for tensor modes @xcite ,
the _ wmap _ data mainly constrain the amplitude of tensor modes by the low-@xmath256 temperature power spectrum ( see section 3.2.3 of * ? ? ?
nevertheless , it is still useful to see how much constraint one can obtain from the 7-year polarization data .
we first fix the cosmological parameters at the 5-year _ wmap _ best - fit values of a power - law @xmath1cdm model .
we then calculate the tensor mode contributions to the b - mode , e - mode , and te power spectra as a function of one parameter : the amplitude , in the form of the tensor - to - scalar ratio , @xmath395 , defined as @xmath414 where @xmath415 is the power spectrum of tensor metric perturbations , @xmath416 , given by @xmath417 in figure [ fig : tens ] , we show the limits on @xmath395 from the b - mode power spectrum only ( @xmath261 , 95% cl ) , from the b- and e - mode power spectra combined ( @xmath263 ) , and from the b - mode , e - mode , and te power spectra combined ( @xmath267 ) .
these limits are significantly better than those from the 5-year data ( @xmath262 , 2.7 , and 1.6 , respectively ) , because of the smaller noise and shifts in the best - fitting values . for comparison ,
the b - mode power spectrum from the bicep 2-year data gives @xmath418 ( 95% cl ; * ? ? ?
if we add the temperature power spectrum , but still fix all the other cosmological parameters including @xmath29 , then we find @xmath419 ( 95% cl ) from both 5-year and 7-year data ; however , due to a strong correlation between @xmath29 and @xmath395 , this would be an underestimate of the error . for a 7-parameter model ( a flat @xmath1cdm model with a tilted spectrum , tensor modes , and @xmath420 ) , we find @xmath421 from the _ wmap _ data alone @xcite , @xmath422 from _ wmap _ plus acbar and quad , @xmath423 from _
wmap_+bao+@xmath0 , and @xmath424 from _ wmap_+bao+sn , where `` sn '' is the constitution samples compiled by @xcite ( see section [ sec : sn ] ) .
we give a summary of these numbers in table [ tab : ns ] .
let us relax the assumption that the power spectrum is a pure power - law , and add a `` running index , '' @xmath388 as @xcite @xmath425 ignoring tensor modes again , we find @xmath426 from _ wmap_+bao+@xmath0 . for comparison , the _ wmap _ data - only limit is @xmath427 @xcite , and the _ wmap_+acbar+quad limit is @xmath428 .
none of these data combinations require @xmath388 : improvements in a goodness - of - fit relative to a power - law model ( equation ( [ eq : powerlaw ] ) ) are @xmath429 , @xmath430 , and @xmath431 for the _ wmap_-only , _ wmap_+acbar+quad , and _ wmap_+bao+@xmath0 , respectively .
see table [ tab : ns ] for the case where both @xmath395 and @xmath388 are allowed to vary .
a simple power - law primordial power spectrum without tensor modes continues to be an excellent fit to the data .
while we have not done a non - parametric study of the shape of the power spectrum , recent studies after the 5-year data release continue to show that there is no convincing deviation from a simple power - law spectrum @xcite .
while the _ wmap _ data alone can not constrain the spatial curvature parameter of the observable universe , @xmath97 , very well , combining the _ wmap _ data with other distance indicators such as @xmath0 , bao , or supernovae can constrain @xmath97 ( e.g. , * ? ? ?
assuming a @xmath1cdm model ( @xmath66 ) , we find @xmath432 from _ wmap_+bao+@xmath0 .. ] however , the limit weakens significantly if dark energy is allowed to be dynamical , @xmath433 , as this data combination , _
wmap_+bao+@xmath0 , can not constrain @xmath89 very well .
we need additional information from type ia supernovae to constrain @xmath89 and @xmath97 simultaneously ( see section 5.3 of * ? ? ?
we shall explore this possibility in section [ sec : darkenergy ] .
non - adiabatic fluctuations are a powerful probe of the origin of matter and the physics of inflation .
following section 3.6 of @xcite , we focus on two physically motivated models for non - adiabatic fluctuations : axion - type @xcite and curvaton - type @xcite . for both cases , we consider non - adiabatic fluctuations between photons and cold dark matter : @xmath434 and report limits on the ratio of the power spectrum of @xmath435 and that of the curvature perturbation @xmath436 ( e.g. , * ? ? ? * ) : @xmath437 where @xmath410 .
we denote the limits on axion - type and curvaton - type by @xmath438 and @xmath439 , respectively .
can also be converted into the numbers showing `` how much the adiabatic relation ( @xmath440 ) can be violated , '' @xmath441 , which can be calculated from @xmath442 } \approx \frac{\sqrt{\alpha}}{3},\ ] ] for @xmath443 @xcite . ]
we find no evidence for non - adiabatic fluctuations .
the _ wmap _
data - only limits are @xmath444 and @xmath445 ( 95% cl ; * ? ? ?
* ) . with _
wmap_+bao+@xmath0 , we find @xmath446 while with _
wmap_+bao+sn , we find @xmath447 and @xmath448 .
the limit on @xmath438 has an important implication for axion dark matter .
in particular , a limit on @xmath438 is related to a limit on the tensor - to - scalar ratio , @xmath395 @xcite .
the explicit formula is given by equation ( 48 ) of @xcite as .
] @xmath449 where @xmath450 is the axion density parameter , @xmath451 is the phase of the peccei - quinn field within our observable universe , and @xmath452 is a `` dilution factor '' representing the amount by which the axion density parameter , @xmath453 , would have been diluted due to a potential late - time entropy production by , e.g. , decay of some ( unspecified ) heavy particles , between 200 mev and the epoch of nucleosynthesis , 1 mev .
where does this formula come from ? within the context of the `` misalignment '' scenario of axion dark matter @xcite . that the peccei - quinn symmetry was not restored after reheating
requires the reheating temperature after inflation not to exceed @xmath454 .
] , there are two observables one can use to place limits on the axion properties : the dark matter density and @xmath438 .
they are given by ( e.g. , * ? ? ?
* and references therein ) @xmath455 where @xmath454 is the axion decay constant , and @xmath456 is the so - called slow - roll parameter ( where @xmath457 is the hubble expansion rate during inflation ) . for single - field inflation models , @xmath458 is related to @xmath395 as @xmath459 . by eliminating the axion decay constant , one obtains equation ( [ eq : rlimitfromaxion ] ) . in deriving the above formula for @xmath453 ( equation ( [ eq : oah2 ] ) ) , we have assumed that the axion field began to oscillate before the qcd phase transition .
, can be calculated from the condition @xmath460 , where @xmath461 is the mass of axions before the qcd phase transition , @xmath462 , and @xmath463 is the mass of axions at the zero temperature . here , we have used the pion decay constant of @xmath464 to calculate @xmath465 , following equation ( 3.4.16 ) of @xcite .
the hubble expansion rate during radiation era is given by @xmath466 , where @xmath467 gev is the reduced planck mass and @xmath468 is the number of relativistic degrees of freedom . before the qcd phase transition , @xmath469 .
after the qcd phase transition but before the electron - positron annihilation , @xmath470 .
] this is true when @xmath471 ; however , when @xmath472 , the axions are so light that the axion field would not start oscillating after the qcd phase transition .
, can be found from the condition @xmath473 and @xmath460 .
see @xcite for more accurate numerical estimate .
note that @xcite used @xmath474 for the pion decay constant when calculating the axion mass at the zero temperature .
] in this limit , the formula for @xmath453 is given by @xmath475 by eliminating @xmath454 from equation ( [ eq : alpha0axion ] ) and ( [ eq : oah22 ] ) , we obtain another formula for @xmath395 : @xmath476 equation ( [ eq : rlimitfromaxion ] ) and ( [ eq : rlimitfromaxion2 ] ) , combined with our limits on @xmath23 and @xmath438 , implies that the axion dark matter scenario in which axions account for most of the observed amount of dark matter , @xmath477 , must satisfy @xmath478 alternatively , one can express this constraint as @xmath479 therefore , a future detection of tensor modes at the level of @xmath480 would imply a fine - tuning of @xmath451 or @xmath481 or both of these parameters @xcite . if such fine - tunings are not permitted , axions can not account for the observed abundance of dark matter ( in the misalignment scenario that we have considered here ) .
depending on one s interest , one may wish to eliminate the phase , leaving the axion decay constant in the formula ( see equation ( b7 ) of * ? ? ? * ) : @xmath482 for @xmath483 .
this formula gives @xmath484 which is inconsistent with the condition @xmath471 ( unless @xmath395 is extremely small ) .
the formula that is valid for @xmath485 is @xmath486 which gives @xmath487 requiring @xmath488 gev , we obtain @xmath489 thus , a future detection of tensor modes at the level of @xmath480 implies a significant amount of entropy production , @xmath490 , or a super - planckian axion decay constant , @xmath491 , or both .
also see @xcite for similar studies . for the implications of @xmath439 for curvaton dark matter , see section 3.6.4 of @xcite .
while the tb and eb correlations vanish in a parity - conserving universe , they may not vanish when global parity symmetry is broken on cosmological scales @xcite . in pixel space
, they would show up as a non - vanishing @xmath492 .
as we showed already in section [ sec : chi2 ] , the _
wmap_7-year @xmath492 data are consistent with noise .
what can we learn from this ?
it is now a routine work of cmb experiments to deliver the tb and eb data , and constrain a rotation angle of the polarization plane due to a parity - violating effect ( or a rotation due to some systematic error ) .
specifically , a rotation of the polarization plane by an angle @xmath284 gives the following 5 transformations : @xmath493 where @xmath494 s on the right hand side are the primordial power spectra in the absence of rotation , while @xmath495 s on the left hand side are what we would observe in the presence of rotation .
note that these equations are not exact but valid only when the primordial @xmath10-mode polarization is negligible compared to the @xmath8-mode polarization , i.e. , @xmath496 . for the full expression including @xmath497 , see @xcite and @xcite . roughly speaking ,
when the polarization data are still dominated by noise rather than by the cosmic signal , the uncertainty in @xmath284 is given by a half of the inverse of the signal - to - noise ratio of te or ee , i.e. , @xmath498 \nonumber & \simeq & \frac1{2(s / n)^{\rm te}},\\ \nonumber { \rm err}[\delta\alpha^{\rm eb } ] & \simeq & \frac1{2(s / n)^{\rm ee}}.\end{aligned}\ ] ] ( note that we use the full likelihood code to find the best - fitting values and error bars . these equations should only be used to provide an intuitive feel of how the errors scale with signal - to - noise . ) as we mentioned in the last paragraph of section [ sec : chi2 ] , with the 7-year polarization data we detect the te power spectrum at @xmath499 from @xmath247 to @xmath248 .
we thus expect @xmath500\simeq 1/42\simeq 0.024~{\rm radian}\simeq 1.4^\circ$ ] , which is significantly better than the 5-year value , @xmath501 @xcite . on the other hand
, we detect the ee power spectrum at @xmath502 only at a few @xmath9 level , and thus @xmath503\gg { \rm err}[\delta\alpha^{\rm tb}]$ ] , implying that we may ignore the high-@xmath256 eb data .
the magnitude of polarization rotation angle , @xmath284 , depends on the path length over which photons experienced a parity - violating interaction .
as pointed out by @xcite , this leads to the polarization angle that depends on @xmath256 .
we can divide this @xmath256-dependence in two regimes : ( i ) @xmath504 : the polarization signal was generated during reionization @xcite .
we are sensitive only to the polarization rotation between the reionization epoch and present epoch .
( ii ) @xmath505 : the polarization signal was generated at the decoupling epoch .
we are sensitive to the polarization rotation between the decoupling epoch and present epoch ; thus , we have the largest path length in this case . using the high-@xmath256 tb data from @xmath247 to @xmath248 , we find @xmath257 , which is a significant improvement over the 5-year high-@xmath256 result , @xmath258 @xcite .
let us turn our attention to lower multipoles , @xmath506 . here
, with the 7-year polarization data , the ee power spectrum is detected at @xmath507 , whereas the te power spectrum is only marginally seen ( @xmath508 ) .
( the overall significance level of detection of the e - model polarization at @xmath506 , including ee and te , is @xmath509 . )
we therefore use both the tb and eb data at @xmath506 . we find @xmath265 , which is also a good improvement over the 5-year low-@xmath256 value , @xmath510 .
combining the low-@xmath256 tb / eb and high-@xmath256 tb data , we find @xmath268 ( the 5-year combined limit was @xmath269 ) , where the quoted error is purely statistical ; however , the _ wmap _ instrument can measure the polarization angle to within @xmath511 of the design orientation @xcite .
we thus add @xmath512 as an estimate of a potential systematic error .
our final 7-year limit is @xmath513 or @xmath75 ( 95% cl ) , for which we have added the statistical and systematic errors in quadrature ( which may be an under - estimate of the total error ) .
the statistical error and systematic error are now comparable .
several research groups have obtained limits on @xmath284 from various data sets @xcite .
recently , the boomerang collaboration @xcite revisited a limit on @xmath284 from their 2003 flight ( b2k ) , taking into account the effect of systematic errors rotating the polarization angle by @xmath514 . by removing this
, they find @xmath515 ( 68% cl ) .
the quad collaboration used their final data set to find @xmath516 ( 68% cl ; * ? ? ? * ) .
@xcite used the bicep 2-year data @xcite to find @xmath517 ( 68% cl statistical ) ; however , a systematic error of @xmath518 needs to be added to this error budget ( see polarization orientation uncertainty in table 3 of * ? ? ?
. therefore , basically the systematic errors in recent measurements of @xmath284 from _ wmap _ 7-year ,
quad final , and bicep 2-year data are comparable to the statistical errors . adding the statistical and systematic errors in quadrature and averaging over _ wmap _ , quad and bicep with the inverse variance weighting
, we find @xmath519 ( 68% cl ) , or @xmath520 ( 95% cl ) .
we therefore conclude that the microwave background data are comfortably consistent with a parity - conserving universe .
see , e.g. , @xcite and references therein for implications of this result for potential violations of lorentz invariance and cpt symmetry . following section 6.1 of @xcite ( also see references therein ) , we constrain the total mass of neutrinos , @xmath521 , mainly from the 7-year _ wmap _ data combined with the distance information .
a new component in the analysis is the exact treatment of massive neutrinos when calculating the likelihood of the bao data , as described in section [ sec : exactnu ] ( also see * ? ? ?
* ) . for a flat @xmath1cdm model , i.e. , @xmath66 and @xmath88 , the _
wmap_-only limit is @xmath522 , while the _ wmap_+bao+@xmath0 limit is @xmath523 the latter is the best upper limit on @xmath322 without information on the growth of structure , which is achieved by a better measurement of the early integrated sachs - wolfe ( isw ) effect through the third acoustic peak of the 7-year temperature power spectrum @xcite , as well as by a better determination of @xmath0 from @xcite . for explanations of this effect ,
see @xcite or section 6.1.3 of @xcite .
@xcite combined the 5-year version of _ wmap_+bao+@xmath0 with the small - scale cmb data to find @xmath524 ev ( 95% cl ) .
therefore , the improvement from this value to our 7-year limit , @xmath78 , indeed comes from a better determination of the amplitude of the third acoustic peak in the 7-year temperature data .
the limit improves when information on the growth of structure is added .
for example , with _ wmap_+@xmath0 and the power spectrum of lrgs ( * ? ?
* see section [ sec : lrg ] ) combined , we find @xmath525 for @xmath66 .
the _ wmap_+bao+@xmath0 limit on the neutrino mass weakens significantly to @xmath522 for @xmath433 because we do not use information of type ia supernovae here to constrain @xmath89 .
this is driven by @xmath89 being too negative : there is an anti - correlation between @xmath89 and @xmath322 @xcite .
the best - fitting value of @xmath89 in this case is @xmath526 ( 68% cl ) .
are anti - correlated implies that the neutrino mass limit would improve if we impose a prior on @xmath89 as @xmath527 . ] for _
wmap_+lrg+@xmath0 , we find @xmath528 ( 95% cl ) for @xmath433 . when the constitution supernova data are included ( _ wmap_+bao+sn ) , we find @xmath529 is slightly weaker than the 5-year _
wmap_+bao+sn limit , 0.67 ev .
the 5-year limit was derived using an approximate treatment of the effect of massive neutrinos on @xmath301 .
the 7-year limit we quote here , which uses the exact treatment of massive neutrinos ( section [ sec : exactnu ] ) , is more reliable . ] and 0.91 ev ( 95% cl ) for @xmath66 and @xmath433 , respectively .
recent studies after the 5-year data release combined the _
wmap_5-year data with information on the growth of structure to find various improved limits .
@xcite added the abundance of x - ray - selected clusters of galaxies , which were found in the _ rosat _ all sky survey and followed up by the _
chandra _ x - ray observatory ( their cluster catalog is described in * ? ? ?
* ) , to the _ wmap _ 5-year data , the bao measurement from @xcite , and the type ia supernova data from @xcite , to find @xmath530 ev ( 95% cl ) for @xmath433 .
@xcite added a different cluster catalog , also derived from the _ rosat _ all sky survey and followed up by the _
chandra _ x - ray observatory ( their cluster catalog is described in * ? ? ?
* ) , and the measurement of the gas mass fraction of relaxed clusters @xcite to the _ wmap _ 5-year data , the bao measurement from @xcite , and the `` union '' type ia supernova samples from @xcite ( all of which constitute the 5-year _ _ wmap__+bao+sn set in * ? ? ?
* ) , to find @xmath530 and 0.43 ev ( 95% cl ) for @xmath66 and @xmath433 , respectively .
@xcite added a prior on the amplitude of matter density fluctuations , @xmath531 ( 68% cl ; * ? ? ? * ) , which was derived from the abundance of optically - selected clusters of galaxies called the `` maxbcg cluster catalog '' @xcite , to the 5-year _
wmap_+bao+sn , and found @xmath532 and 0.52 ev ( 95% cl ) for @xmath66 and @xmath433 , respectively .
@xcite added the angular power spectra of photometrically selected samples of lrgs called `` megaz '' to the 5-year _
wmap_+bao+sn , and found @xmath533 ev ( 95% cl ) for @xmath66 .
@xcite pointed out that the limit on @xmath534 from galaxy clusters would improve significantly by not only using the abundance but also the power spectrum of clusters . in order to exploit the full information contained in the growth of structure , it is essential to understand the effects of massive neutrinos on the _ non - linear _ growth . all of the work to date ( including _ wmap_+lrg+@xmath0 presented above ) included the effects of massive neutrinos on the _ linear _ growth , while ignoring their non - linear effects
. the widely - used phenomenological calculation of the non - linear matter power spectrum called the halofit @xcite has not been calibrated for models with massive neutrinos .
consistent treatments of massive neutrinos in the non - linear structure formation using cosmological perturbation theory @xcite and numerical simulations @xcite have just begun to be explored .
more work along these lines would be necessary to exploit the information on the growth structure to constrain the mass of neutrinos .
lccccc @xmath535 & 5-year & @xmath536 & @xmath537 & & & 7-year & @xmath538 & & @xmath539 & @xmath540 @xmath52 & 5-year & @xmath541 & @xmath542 & & & 7-year & @xmath543 & & @xmath544 & @xmath545 @xmath323 & 5-year & @xmath546 ( 95% cl ) & @xmath547 & & & 7-year & @xmath548 & & @xmath83 & @xmath549 [ tab : neff ] how many relativistic species are there in the universe after the matter - radiation equality epoch ? we parametrize the relativistic degrees of freedom using the effective number of neutrino species , @xmath323 , given in equation ( [ eq : neff ] ) .
this quantity can be written in terms of the matter density , @xmath52 , and the redshift of matter - radiation equality , @xmath535 , as ( see equation ( 84 ) of * ? ? ?
* ) @xmath550 ( here , @xmath551 and @xmath552 are the 5-year maximum likelihood values from the simplest @xmath1cdm model . ) this formula suggests that the variation in @xmath323 is given by @xmath553 the equality redshift is one of the direct observables from the temperature power spectrum .
the _ wmap _
data constrain @xmath535 mainly from the ratio of the first peak to the third peak . as the 7-year temperature power spectrum has a better determination of the amplitude of the third peak @xcite
, we expect a better limit on @xmath535 compared to the 5-year one . for models where @xmath323 is different from 3.04
, we find @xmath554 ( 68% cl ) from the _ wmap _ data only , we find @xmath555 ( 68% cl ) . ] , which is better than the 5-year limit by more than 10% ( see table [ tab : neff ] ) .
however , the fractional error in @xmath52 is much larger , and thus we need to determine @xmath52 using external data .
the bao data provide one constraint .
we also find that @xmath52 and @xmath0 are strongly correlated in the models with @xmath556 ( see figure [ fig : omh2_vs_h ] ) .
therefore , an improved measurement of @xmath0 from @xcite would help reduce the error in @xmath52 , thereby reducing the error in @xmath323 .
the limit on @xmath52 from the 7-year _
wmap_+bao+@xmath0 combination is better than the 5-year `` _ wmap_+bao+sn+_hst _ '' limit by 36% .
we find that the _ wmap_+bao+@xmath0 limit on @xmath323 is @xmath557 while the _ wmap_+lrg+@xmath0 limit is @xmath558 ( 68% cl ) , which are significantly better than the 5-year _ wmap_+bao+sn+_hst
_ limit , @xmath559 ( 68% cl ) .
@xcite added the maxbcg prior , @xmath531 ( 68% cl ; * ? ? ? * ) , to the 5-year _
wmap_+bao+sn+_hst _ , and found @xmath560 ( 68% cl ) .
they also added the above prior to the 5-year version of _
wmap_+lrg+@xmath0 , finding @xmath561 ( 68% cl ) .
the constraint on @xmath323 can also be interpreted as an upper bound on the energy density in primordial gravitational waves with frequencies @xmath562 hz .
many cosmological mechanisms for the generation of stochastic gravitational waves exist , such as certain inflationary models , electroweak phase transitions , and cosmic strings . at low frequencies ( @xmath563 hz ) ,
the background is constrained by the limit on tensor fluctuations described in section [ sec : gw ] .
constraints at higher frequencies come from pulsar timing measurements at @xmath564 hz @xcite , recent data from the laser interferometer gravitational wave observatory ( ligo ) at 100 hz ( with limits of @xmath565 * ? ? ?
* ) , and at frequencies @xmath566 hz from measurements of light - element abundances .
a large gravitational wave energy density at nucleosynthesis would alter the predicted abundances , and observations imply an upper bound of @xmath567 @xcite .
the cmb provides a limit that reaches down to @xmath568 hz , corresponding to the comoving horizon at recombination .
the gravitational wave background within the horizon behaves as free - streaming massless particles , so affects the cmb and matter power spectra in the same way as massless neutrinos @xcite .
the density contributed by @xmath569 massless neutrino species is @xmath570 .
constraints have been found using the wmap 3-year data combined with additional cosmological probes by @xcite , for both adiabatic and homogeneous initial conditions for the tensor perturbations . with the current wmap+bao+@xmath0 data combination ,
we define @xmath571 , and find limits of @xmath572 for adiabatic initial conditions , imposing an @xmath573 prior .
adiabatic conditions might be expected if the gravitational waves were generated by the appearance of cusps in cosmic strings @xcite . for the wmap+lrg+@xmath0 data ,
we find @xmath574 , or @xmath575 at 95% cl .
given a particular string model , these bounds can be used to constrain the cosmic string tension ( e.g. , * ? ? ?
* ; * ? ? ?
* ) .
a change in the primordial helium abundance affects the shape of the temperature power spectrum @xcite .
the most dominant effect is a suppression of the power spectrum at @xmath576 due to an enhanced silk damping effect . for a given mass density of baryons ( protons and helium nuclei )
, the number density of electrons , @xmath577 , can be related to the primordial helium abundance .
when both hydrogen and helium were ionized , @xmath578 .
however , most of the helium recombines by @xmath579 ( see * ? ? ?
* and references therein ) , much earlier than the photon decoupling epoch , @xmath7 . as a result , the number density of free electrons at around the decoupling epoch is given by @xmath580 @xcite .
the larger @xmath17 is , the smaller @xmath577 becomes .
if the number of electrons is reduced , photons can free - stream longer ( the mean free path of photons , @xmath581 , gets larger ) , wiping out more temperature anisotropy .
therefore , a larger @xmath17 results in a greater suppression of power on small angular scales .
@xcite ( also see * ? ? ?
* ) show that a 100% change in @xmath17 changes the heights of the second , third , and forth peaks by @xmath582% , 3% , and 3% , respectively .
therefore , one expects that a combination of the _ wmap _ data and small - scale cmb experiments such as acbar @xcite and quad @xcite would be a powerful probe of the primordial helium abundance . in figure [
fig : zerohelium ] , we compare the _ wmap _ , acbar , and quad data with the temperature power spectrum with the nominal value of the primordial helium abundance , @xmath583 ( pink line ) , and that with a tiny amount of helium , @xmath584 ( blue line ) . there is too much power in the case of @xmath584 , making it possible to _ detect _ the primordial helium effect using the cmb data alone .
however , one must be careful about a potential degeneracy between the effect of helium and those of the other cosmological parameters .
first , as the number density of electrons is given by @xmath585 , @xmath17 and @xmath586 may be correlated .
second , a scale - dependent suppression of power such as this may be correlated with the effect of tilt , @xmath29 @xcite . in the left panel of figure [ fig : obh2_vs_yhe ]
, we show that @xmath586 and @xmath17 are essentially uncorrelated : the baryon density is determined by the first - to - second peak ratio relative to the first - to - third peak ratio , which is now well measured by the
_ wmap_data .
therefore , the current _ wmap _ data allow @xmath586 to be measured regardless of @xmath17 . in the middle panel of figure [ fig : obh2_vs_yhe ]
, we show that there is a slight positive correlation between @xmath29 and @xmath17 : an enhanced silk damping produced by a larger @xmath17 can be partially canceled by a larger @xmath29 @xcite .
we find a 95% cl upper limit of @xmath587 from the _ wmap _ data alone .
when we add the acbar and quad data , _ we find a significant detection of the effect of primordial helium by more than @xmath588 _
( see the right panel of figure [ fig : obh2_vs_yhe ] ) , @xmath589 the 95% cl limit is @xmath590 .
the 99% cl lower limit is @xmath591 .
this value is broadly consistent with the helium abundances estimated from observations of low - metallicity extragalactic ionized ( hii ) regions , @xmath592 @xcite .
see @xcite for a review .
we can improve this limit by imposing an _ upper limit _ on @xmath17 from these astrophysical measurements .
as the helium is created by nuclear fusion in stars , the helium abundances measured from stars ( e.g. , sun ; see * ? ? ? * for a recent review ) and hii regions are , in general , larger than the primordial abundance . on the other hand ,
as we have just shown , the cmb data provide a _ lower limit _ on @xmath17 .
even with a very conservative hard prior , @xmath593 , we find @xmath594 ( 68% cl ) , is found such that the integral of the posterior likelihood of @xmath17 in @xmath595 is 68% of the integral in @xmath596 .
similarly , the 95% cl lower limit is @xmath597 and the 99% cl lower limit is @xmath598 . ] .
therefore , a combination of the cmb and the solar constraints on @xmath17 offers a new way for testing the predictions of theory of the big bang nucleosynthesis ( bbn ) .
for example , the bbn predicts that the helium abundance is related to the baryon - to - photon ratio , @xmath599 , and the number of additional neutrino species ( or any other additional relativistic degrees of freedom ) during the bbn epoch , @xmath600 , as ( see equation ( 11 ) of * ? ? ? * ) @xmath601,\ ] ] where @xmath602 and @xmath603 ( 68% cl ; _
wmap_+bao+@xmath0 ) .
( see * ? ? ? * for more discussion on this method . ) for @xmath604 , the helium abundance changes by @xmath605 , which is much smaller than our error bar , but is comparable to the expected error bar from _ planck _
@xcite .
there have been several attempts to measure @xmath17 from the cmb data @xcite .
the previous best - limit is @xmath606 at 68% cl ( 95% cl ) , which was obtained by @xcite from the _ wmap _ 5-year data combined with acbar @xcite , boomerang @xcite , and cosmic background imager ( cbi ; * ? ? ?
note that the likelihood function of @xmath17 is non - gaussian , with a tail extending to @xmath607 ; thus , the level of significance of detection was less than @xmath588 .
in this section , we provide limits on the properties of dark energy , characterized by the equation of state parameter , @xmath89 .
we first focus on constant ( time independent ) equation of state in a flat universe ( section [ sec : w_flat ] ) and a curved universe ( section [ sec : w_curve ] ) .
we then constrain a time - dependent @xmath89 given by @xmath608 , where @xmath609 is the scale factor , in section [ sec : w0wa ] .
next , we provide the 7-year `` _ wmap _ normalization prior '' in section [ sec : normalization_prior ] , which is useful for constraining @xmath89 ( as well as the mass of neutrinos ) from the growth of cosmic density fluctuations .
( see , e.g. , * ? ? ? * for an application of the 5-year normalization prior to the x - ray cluster abundance data . ) in section [ sec : distance_prior ] , we provide the 7-year `` _ wmap _ distance prior , '' which is useful for constraining a variety of time - dependent @xmath89 models for which the markov chain monte carlo exploration of the parameter space may not be available .
( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for applications of the 5-year distance prior . )
we give a summary of our limits on dark energy parameters in table [ tab : darkenergy ] . in a flat universe , @xmath88
, an accurate determination of @xmath0 helps improve a limit on a constant equation of state , @xmath89 @xcite . using _
wmap_+bao+@xmath0 , we find @xmath610 which improves to @xmath611 ( 68% cl ) if we add the time - delay distance out to the lens system b1608 + 656 ( * ? ? ?
* see section [ sec : timedelay ] ) . _ these limits are independent of high-@xmath294 type ia supernova data .
_ the high-@xmath294 supernova data provide the most stringent limit on @xmath89 . using _
wmap_+bao+sn , we find @xmath612 ( 68% cl ) .
the error does not include systematic errors in supernovae , which are comparable to the statistical error @xcite ; thus , the error in @xmath89 from _ wmap_+bao+sn is about a half of that from _ wmap_+bao+@xmath0 or _ wmap_+bao+@xmath0+@xmath101 .
the cluster abundance data are sensitive to @xmath89 via the comoving volume element , angular diameter distance , and growth of matter density fluctuations @xcite . by combining the cluster abundance data and
the 5-year _ wmap _ data , @xcite found @xmath613 ( 68% cl ) for a flat universe . by adding bao of @xcite and the supernova data of @xcite
, they found @xmath614 ( 68% cl ) .
these results using the cluster abundance data ( also see * ? ? ?
* ) agree well with our corresponding _
wmap_+bao+@xmath0 and _ wmap_+bao+sn limits . , and the curvature parameter , @xmath615 .
the contours show the 68% and 95% cl from _
wmap_+bao+@xmath0 ( red ) , _
wmap_+bao+@xmath0+@xmath101 ( black ) , and _ wmap_+bao+sn ( purple ) . , width=302 ] when @xmath616 , limits on @xmath89 significantly weaken , with a tail extending to large negative values of @xmath89 , unless supernova data are added . in figure
[ fig : wok ] , we show that _
wmap_+bao+@xmath0 allows for @xmath617 , which can be excluded by adding information on the time - delay distance . in both cases ,
the spatial curvature is well constrained : we find @xmath618 from _
wmap_+bao+@xmath0 , and @xmath619 ( 68% cl ) from _
wmap_+bao+@xmath0+@xmath620 , whose errors are comparable to that of the _ wmap_+bao+@xmath0 limit on @xmath97 with @xmath66 , @xmath621 ( 68% cl ; see section [ sec : ok ] ) .
however , @xmath89 is poorly constrained : we find @xmath622 from _
wmap_+bao+@xmath0 , and @xmath623 ( 68% cl ) from _ wmap_+bao+@xmath0+@xmath101 . among the data combinations that do not use the information on the growth of structure ,
the most powerful combination for constraining @xmath97 and @xmath89 simultaneously is a combination of the _ wmap _ data , bao ( or @xmath101 ) , and supernovae , as _ wmap_+bao ( or @xmath101 ) primarily constrains @xmath97 , and _
wmap_+sn primarily constrains @xmath89 . with _
wmap_+bao+sn , we find @xmath624 and @xmath625 ( 68% cl ) .
note that the error in the curvature is essentially the same as that from _
wmap_+bao+@xmath0 , while the error in @xmath89 is @xmath626 times smaller .
@xcite combined their cluster abundance data with the 5-year _
wmap_+bao+sn to find @xmath627 ( 68% cl ) for a curved universe .
@xcite combined their lrg power spectrum with the 5-year _ wmap _ data and the union supernova data to find @xmath628 and @xmath629 ( 68% cl ) .
these results are in good agreement with our 7-year _
wmap_+bao+sn limit . .
the contours show the 68% and 95% cl from _
wmap_+@xmath0+sn ( red ) , _
wmap_+bao+@xmath0+sn ( blue ) , and _ wmap_+bao+@xmath0+@xmath101+sn ( black ) , for a flat universe . , width=302 ] as for a time - dependent equation of state , we shall find constraints on the present - day value of the equation of state and its derivative using a linear form , @xmath608 @xcite . we assume a flat universe , @xmath88 .
( for recent limits on @xmath382 with @xmath616 , see * ? ? ? * and references therein . ) while we have constrained this model using the _ wmap _ distance prior in the 5-year analysis ( see section 5.4.2 of * ? ? ?
* ) , in the 7-year analysis we shall present the full markov chain monte carlo exploration of this model . for a time - dependent equation of state , one must be careful about the treatment of perturbations in dark energy when @xmath89 crosses @xmath630 .
we use the `` parametrized post - friedmann '' ( ppf ) approach , implemented in the camb code following @xcite . crossing @xmath630 .
the constraints on @xmath102 and @xmath106 have been obtained using this model and the previous years of _ wmap _ data @xcite . ] in figure [ fig : w0wa ] , we show the 7-year constraints on @xmath102 and @xmath106 from _
wmap_+@xmath0+sn ( red ) , _
wmap_+bao+@xmath0+sn ( blue ) , and _ wmap_+bao+@xmath0+@xmath101+sn ( black ) . the angular diameter distances measured from bao and @xmath101 help exclude models with large negative values of @xmath106 .
we find that the current data are consistent with a cosmological constant , even when @xmath89 is allowed to depend on time .
however , a large range of values of @xmath631 are still allowed by the data : we find @xmath632 from _ wmap_+bao+@xmath0+sn . when the time - delay distance information is added , the limits improve to @xmath633 and @xmath634 ( 68% cl ) .
@xcite combined their cluster abundance data with the 5-year _
wmap_+bao+sn to find a limit on a linear combination of the parameters , @xmath635 ( 68% cl ) .
our data combination is sensitive to a different linear combination : we find @xmath636 ( 68% cl ) for the 7-year _
wmap_+bao+@xmath0+sn combination .
the current data are consistent with a flat universe dominated by a cosmological constant .
the growth of cosmological density fluctuations is a powerful probe of dark energy , modified gravity , and massive neutrinos .
the _ wmap _
data provide a useful normalization of the cosmological perturbation at the decoupling epoch , @xmath7 . by comparing this normalization with the amplitude of matter density fluctuations in a low redshift universe
, one may distinguish between dark energy and modified gravity @xcite and determine the mass of neutrinos @xcite . in section 5.5 of @xcite
, we provided a `` _ wmap _ normalization prior , '' which is a constraint on the power spectrum of curvature perturbation , @xmath637 .
@xcite combined this with the number density of clusters of galaxies to constrain the dark energy equation of state , @xmath89 , and the amplitude of matter density fluctuations , @xmath39 .
the matter density fluctuation in fourier space , @xmath638 , is related to @xmath639 as @xmath640 , where @xmath641 and @xmath642 are the linear growth rate and the matter transfer function normalized such that @xmath643 as @xmath644 , and @xmath645 as @xmath644 during the matter era , respectively . ignoring the mass of neutrinos and modifications to gravity
, one can obtain the growth rate by solving a single differential equation @xcite .
needs to be replaced by @xmath382 . ]
the 7-year normalization prior is @xmath646 where @xmath647 . for comparison ,
the 5-year normalization prior was @xmath648 .
this normalization prior is valid for models with @xmath93 , @xmath649 , or @xmath650 .
however , these normalizations can not be used for the models that have the tensor modes , @xmath651 , or the running index , @xmath652 .
lccc @xmath653 & 302.09 & 302.69 & 0.76
@xmath654 & 1.725 & 1.726 & 0.018 @xmath655 & 1091.3 & 1091.36 & 0.91 [ tab : wmap_prior ] lccc @xmath653 & @xmath656 & @xmath657 & @xmath658 @xmath654 & & @xmath659 & @xmath660 @xmath655 & & & @xmath661 [ tab : wmap_prior_cov ] the temperature power spectrum of cmb is sensitive to the physics at the decoupling epoch , @xmath7 , as well as the physics between now and the decoupling epoch .
the former primarily affects the amplitude of acoustic peaks , i.e. , the ratios of the peak heights and the silk damping .
the latter changes the _ locations _ of peaks via the angular diameter distance out to the decoupling epoch .
one can quantify this by ( i ) the `` acoustic scale '' , @xmath653 , @xmath662 where @xmath655 is the redshift of decoupling , for which we use the fitting formula of @xcite , as well as by ( ii ) the `` shift parameter , '' @xmath654 @xcite , @xmath663 these two parameters capture most of the constraining power of the _
wmap_data for dark energy properties @xcite , with one important difference .
the distance prior does not capture the information on the growth of structure probed by the late - time isw effect .
as a result , the dark energy constraints derived from the distance prior are similar to , but weaker than , those derived from the full analysis @xcite .
we must understand the limitation of this method .
namely , the distance prior is applicable only when the model in question is based on : * the standard friedmann - lemaitre - robertson - walker universe with matter , radiation , dark energy , as well as spatial curvature , * neutrinos with the effective number of neutrinos equal to 3.04 , and the minimal mass ( @xmath664 ev ) , and * nearly power - law primordial power spectrum of curvature perturbations , @xmath665 , negligible primordial gravitational waves relative to the curvature perturbations , @xmath666 , and negligible entropy fluctuations relative to the curvature perturbations , @xmath667 .
in table [ tab : wmap_prior ] and [ tab : wmap_prior_cov ] , we provide the 7-year distance prior . the errors in @xmath653 , @xmath654 , and @xmath655 have improved from the 5-year values by 12% , 5% , and 2% , respectively . to compute the likelihood , use @xmath668 where @xmath669 is the values predicted by a model in question
, @xmath670 is the data given in table [ tab : wmap_prior ] , and @xmath671 is the inverse covariance matrix given in table [ tab : wmap_prior_cov ] .
also see section 5.4.1 of @xcite for more information .
during the period of cosmic inflation @xcite , quantum fluctuations were generated and became the seeds for the cosmic structures that we observe today @xcite .
( also see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for reviews . )
inflation predicts that the statistical distribution of primordial fluctuations is nearly a gaussian distribution with random phases .
measuring deviations from a gaussian distribution , i.e. , _ non - gaussian _ correlations in primordial fluctuations , is a powerful test of inflation , as how precisely the distribution is ( non-)gaussian depends on the detailed physics of inflation ( see * ? ? ?
* ; * ? ? ?
* for reviews ) . in this paper
, we constrain the amplitude of non - gaussian correlations using the angular bispectrum of cmb temperature anisotropy , the harmonic transform of the 3-point correlation function ( see * ? ? ?
* for a review ) .
the observed angular bispectrum is related to the 3-dimensional bispectrum of primordial curvature perturbations , @xmath672 . in the linear order ,
the primordial curvature perturbation is related to bardeen s curvature perturbation @xcite in the matter - dominated era , @xmath673 , by @xmath674 ( e.g. , * ? ? ?
the cmb temperature anisotropy in the sachs
wolfe limit @xcite is given by @xmath675 .
we write the bispectrum of @xmath673 as @xmath676 we shall explore 3 different shapes of the primordial bispectrum : `` local , '' `` equilateral , '' and `` orthogonal . ''
they are defined as follows : * * local form*. the local form bispectrum is given by @xcite @xmath677\\ & = & 2a^2{f_{nl}^{\rm local}}\left[\frac1{k^{4-n_s}_1k^{4-n_s}_2}+(2~\mbox{perm}.)\right ] , \label{eq : flocal}\end{aligned}\ ] ] where @xmath678 is the power spectrum of @xmath673 with a normalization factor @xmath679 .
this form is called the local form , as this bispectrum can arise from the curvature perturbation in the form of @xmath680 , where both sides are evaluated at the same location in space ( @xmath681 is a linear gaussian fluctuation ) .
is not the only way to produce this type of bispectrum .
one can also produce this form from multi - scalar field inflation models where scalar field fluctuations are nearly scale invariant @xcite ; multi - scalar models called `` curvaton '' scenarios @xcite ; multi - field models in which one field modulates the decay rate of inflaton field @xcite ; multi - field models in which a violent production of particles and non - linear reheating , called `` preheating , '' occur due to parametric resonances @xcite ; models in which the universe contracts first and then bounces ( see * ? ? ?
* for a review ) .
] equation ( [ eq : flocal ] ) peaks at the so - called `` squeezed '' triangle for which @xmath682 @xcite . in this limit
, we obtain @xmath683 how large is @xmath684 from inflation ?
the earlier calculations showed that @xmath684 from single - field slow - roll inflation would be of order the slow - roll parameter , @xmath685 @xcite .
more recently , @xcite and @xcite found that the coefficient of @xmath686 from the simplest single - field slow - roll inflation with the canonical kinetic term in the squeezed limit is given by @xmath687 comparing this result with the form predicted by the @xmath684 model , one obtains @xmath688 , which gives @xmath689 for @xmath690 . * * equilateral form*. the equilateral form bispectrum is given by @xcite @xmath691\right\}. \label{eq : fequil}\end{aligned}\ ] ] this function approximates the bispectrum forms that arise from a class of inflation models in which scalar fields have non - canonical kinetic terms .
one example is the so - called dirac - born - infeld ( dbi ) inflation @xcite , which gives @xmath692 in the limit of @xmath693 , where @xmath694 is the effective sound speed at which scalar field fluctuations propagate .
there are various other models that can produce @xmath695 @xcite .
the local and equilateral forms are nearly orthogonal to each other , which means that both can be measured nearly independently . * * orthogonal form*. the orthogonal form , which is constructed such that it is nearly orthogonal to both the local and equilateral forms , is given by @xcite @xmath696\right\}. \label{eq : forthog}\end{aligned}\ ] ] this form approximates the forms that arise from a linear combination of higher - derivative scalar - field interaction terms , each of which yields forms similar to the equilateral shape .
@xcite found that , using the `` effective field theory of inflation '' approach @xcite , a certain linear combination of similarly equilateral shapes can yield a distinct shape which is orthogonal to both the local and equilateral forms .
note that these are not the most general forms one can write down , and there are other forms which would probe different aspects of the physics of inflation @xcite . of these forms
, the local form bispectrum has special significance .
@xcite showed that not only models with the canonical kinetic term , but _ all _ single - inflation models predict the bispectrum in the squeezed limit given by eq .
( [ eq : singleprediction ] ) , regardless of the form of potential , kinetic term , slow - roll , or initial vacuum state ( also see * ? ? ? * ; * ? ? ? * ; * ? ? ?
this means that a convincing detection of @xmath684 would rule out _ all _ single - field inflation models .
lccccc v+w & raw & @xmath697 & @xmath698 & @xmath699 & n / a v+w & clean & @xmath700 & @xmath701 & @xmath702 & n / a v+w & marg . & @xmath703 & @xmath704 & @xmath705 & @xmath706v & marg . & @xmath707 & @xmath708 & @xmath709 & @xmath710 w & marg . & @xmath711 & @xmath712 & @xmath713 & @xmath714 [ tab : fnl ] the first limit on @xmath684 was obtained from the _ cobe _ 4-year data @xcite by @xcite , using the angular bispectrum .
the limit was improved by an order of magnitude when the _ wmap _ first year data were used to constrain @xmath684 @xcite . since then
the limits have improved steadily as _ wmap _
collect more years of data and the bispectrum method for estimating @xmath684 has improved @xcite . , which do not use the bispectrum directly , see section 3.5 of @xcite .
recently , the `` skewness power spectrum '' has been proposed as a new way to measure @xmath684 and other non - gaussian components such as the secondary anisotropies and point sources @xcite . in the limit that noise is uniform ,
their estimator is equivalent to that of @xcite , which also allows for simultaneous estimations of multiple sources of non - gaussianity ( see appendix a of * ? ? ?
the skewness power spectrum method provides a means to visualize the shape of various bispectra as a function of multipoles . ] in this paper , we shall adopt the optimal estimator ( developed by * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , which builds on and significantly improves the original bispectrum estimator proposed by @xcite , especially when the spatial distribution of instrumental noise is not uniform . for details of the method ,
see appendix a of @xcite for @xmath684 , and section 4.1 of @xcite for @xmath695 and @xmath715 . to construct the optimal estimators
, we need to specify the cosmological parameters .
we use the 5-year @xmath1cdm parameters from _ wmap_+bao+sn , for which @xmath716 .
we also constrain the bispectrum due to residual ( unresolved ) point sources , @xmath717 .
the optimal estimator for @xmath717 is constructed by replacing @xmath718 in equation ( a24 ) of @xcite with @xmath719 , and using their equations ( a17 ) and ( a5 ) .
the @xmath720 matrix is computed by the multigrid - based algorithm of @xcite .
we use the v- and w - band maps at the healpix resolution @xmath721 . as the optimal estimator weights the data optimally at all multipoles
, we no longer need to choose the maximum multipole used in the analysis , i.e. , we use all the data .
we use both the raw maps ( before cleaning foreground ) and foreground - reduced ( clean ) maps to quantify the foreground contamination of @xmath19 parameters . for all cases
, we find the best limits on @xmath19 parameters by combining the v- and w - band maps , and marginalizing over the synchrotron , free - free , and dust foreground templates @xcite . as for the mask
, we always use the _ kq75y7 _ mask @xcite . in table
[ tab : fnl ] , we summarize our results : * * local form results*. the 7-year best estimate of @xmath684 is @xmath722 the 95% limit is @xmath84 .
when the raw maps are used , we find @xmath723 ( 68% cl ) .
when the clean maps are used , but foreground templates are not marginalized over , we find @xmath724 ( 68% cl ) .
these results ( in particular the clean - map versus the foreground marginalized ) indicate that the foreground emission makes a difference at the level of @xmath725 . and
@xmath726 ( 68% cl ) with and without the foreground marginalization , respectively . ]
we find that the v+w result is lower than the v - band or w - band results .
this is possible , as the v+w result contains contributions from the cross - correlations of v and w such as @xmath727 and @xmath728 .
* * equilateral form results*. the 7-year best estimate of @xmath695 is @xmath729 the 95% limit is @xmath85 . for @xmath695 , the foreground marginalization does not shift the central values very much , @xmath730
this makes sense , as the equilateral bispectrum does not couple small - scale modes to very large - scale modes @xmath731 , which are sensitive to the foreground emission .
on the other hand , the local form bispectrum is dominated by the squeezed triangles , which do couple large and small scales modes . * * orthogonal form results*. the 7-year best estimate of @xmath715 is @xmath732 the 95% limit is @xmath86 .
the foreground marginalization has little effect , @xmath733 . as for the point - source bispectrum
, we do not detect @xmath717 in v , w , or v+w . in @xcite
, we estimated that the residual sources could bias @xmath684 by a small positive amount , and applied corrections using monte carlo simulations . in this paper , we do not attempt to make such corrections , but we note that sources could give @xmath734 ( note that the simulations used by @xcite likely overestimated the effect of sources by a factor of two ) . as the estimator has changed from that used by @xcite , extrapolating the previous results
is not trivial .
source corrections to @xmath695 and @xmath715 could be larger @xcite , but we have not estimated the magnitude of the effect for the 7-year data .
we used the linear perturbation theory to calculate the angular bispectrum of primordial non - gaussianity @xcite .
second - order effects @xcite are expected to give @xmath735 @xcite and are negligible given the noise level of the _ wmap _
7-year data . among various sources of
secondary non - gaussianities which might contaminate measurements of primordial non - gaussianity ( in particular @xmath684 ) , a coupling between the isw effect and the weak gravitational lensing is the most dominant source of confusion for @xmath684 @xcite .
while this contribution is expected to be detectable and bias the measurement of @xmath684 for _ planck _ , it is expected to be negligible for _ wmap _ : using the method of @xcite , we estimate that the expected signal - to - noise ratio of this term in the _ wmap _
7-year data is about @xmath736 .
we also estimate that this term can give @xmath684 a potential positive bias of @xmath737 .
@xcite used the skewness power spectrum method of @xcite to search for this term in the _ wmap _ 5-year data and found a null result .
if we subtract @xmath738 estimated above ( for the residual source and the isw - lensing coupling ) from the measured value , @xmath738 becomes more consistent with zero . from these results ,
we conclude that the _ wmap _
7-year data are consistent with gaussian primordial fluctuations to within 95% cl . when combined with the limit on @xmath684 from _ sdss _ , @xmath739 ( 95% cl
* ) , we find @xmath740 ( 95% cl ) .
we review the basics of the sz effect in section [ sec : szintro ] . in section [ sec : coma ] , we shall test our optimal estimator for extracting the sz signal from the _ wmap _ data using the brightest sz source on the sky : the coma cluster .
we also present an improved measurement of the sz effect toward the coma cluster ( @xmath741 ) .
the most significant result from section [ sec : nearby ] is the discovery of the thermal / dynamical effect of clusters on the sz effect .
we shall present the measurements of the sz effects toward nearby ( @xmath742 ) galaxy clusters in vikhlinin et al.s sample @xcite , which were used to infer the cosmological parameters @xcite .
we then compare the measured sz flux to the expected flux from the x - ray data on the individual clusters , finding a good agreement .
significance of detection ( from merely 11 clusters , excluding coma ) is @xmath743 . by dividing the sample into cooling - flow and non - cooling - flow clusters ( or relaxed and non - relaxed clusters ) , we find a significant difference in the sz effect between these sub - samples . in section [ sec : szstacking ] , we shall report a significant ( @xmath744 ) statistical detection of the sz effect at hundreds of positions of the known clusters .
we then compare the measured sz flux to theoretical models as well as to an x - ray - calibrated empirical model , and discuss implications of our measurement , especially a recent measurement of the lower - than - theoretically - expected sz power spectrum by the spt collaboration .
note that the analyses presented in section [ sec : nearby ] and [ sec : szstacking ] are similar but different in one important aspect : the former uses a handful ( 29 ) of clusters with well - measured _ chandra _ x - ray data , while the latter uses hundreds of clusters without detailed x - ray data .
therefore , while the latter results have smaller statistical errors ( and much larger systematic errors ) , the former results have much smaller systematic errors ( and larger statistical errors ) . when cmb photons encounter hot electrons in clusters of galaxies , the temperature of cmb changes due to the inverse compton scattering by these electrons .
this effect , known as the thermal sz effect @xcite , is a source of significant additional ( secondary ) anisotropies in the microwave sky ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for reviews )
. the temperature change due to the sz effect in units of thermodynamic temperature , @xmath745 , depends on frequency , @xmath746 , and is given by ( for a spherically symmetric distribution of gas ) : @xmath747 where @xmath127 is the angular distance from the center of a cluster of galaxies on the sky , @xmath748 the proper ( not comoving ) angular diameter distance to the cluster center , @xmath256 the radial coordinates from the cluster center along the line of sight , @xmath749 the _ electron _ pressure profile , @xmath750 the thomson cross section , @xmath751 the electron mass , @xmath752 the speed of light , and @xmath753 the spectral function given by @xmath754 where @xmath755 for @xmath756 k. in the rayleigh - jeans limit , @xmath757 , one finds @xmath758 . at the _ wmap _ frequencies , @xmath759 , @xmath760 , @xmath761 , @xmath762 , and @xmath763 at 23 , 33 , 41 , 61 , and 94 ghz , respectively .
the integration boundary , @xmath764 , will be given later .
the thermal sz effect ( when relativistic corrections are ignored ) vanishes at @xmath765 ghz .
one then finds @xmath766 at higher frequencies ; thus , the thermal sz effect produces a temperature decrement at @xmath767 ghz , vanishes at 217 ghz , and produces a temperature increment at @xmath768 ghz .
the angular power spectrum of temperature anisotropy caused by the sz effect is sensitive to both the gas distribution in clusters @xcite and the amplitude of matter density fluctuations , i.e. , @xmath39 @xcite . while we have not detected the sz _ power spectrum _ in the _ wmap _ data , we have detected the sz signal from the coma cluster ( abell 1656 ) in the 1-year @xcite and 3-year @xcite data .
we have also made a _
statistical _ detection of the sz effect by cross - correlating the _ wmap _ data with the locations of known clusters in the x - ray brightest abell - type cluster ( xbac ; * ? ? ?
* ) catalog @xcite .
in addition , there have been a number of statistical detections of the sz effect reported by many groups using various methods @xcite .
k ) . while the v- ( green ) and w - band ( blue ) measurements are contaminated by the cmb fluctuations around coma , our optimal estimator can separate the sz effect and cmb when the v- and w - band measurements are combined ( red ) . the solid line shows the best - fitting spherical @xmath769 model with the core radius of @xmath770 arcmin and @xmath771 .
the best - fitting central temperature decrement ( fit to a @xmath769 model ) is @xmath772k .
note that 10 arcmin corresponds to the physical distance of 0.195 @xmath773 mpc at the location of coma .
the radius within which the mean overdensity is 500 times the critical density of the universe , @xmath774 , corresponds to about 50 arcmin . , width=302 ] the coma cluster ( abell 1656 ) is a nearby ( @xmath775 ) massive cluster located near the north galactic pole ( @xmath256 , @xmath776)=(@xmath777 , @xmath778 ) .
the angular diameter distance to coma , calculated from @xmath775 and @xmath779 , is @xmath780 mpc ; thus , 10 arcmin on the sky corresponds to the physical distance of @xmath781 mpc at the redshift of coma . to extract the sz signal from the _ wmap _ temperature map
, we use the optimal method described in appendix [ app : sz ] : we write down the likelihood function that contains cmb , noise , and the sz effect , and marginalize it over cmb . from the resulting likelihood function for the sz effect , which is given by equation ( [ eq : sz_p_d ] ) , we find the optimal estimator for the sz effect in a given angular bin @xmath782 , @xmath783 , as @xmath784^{-1}_{\nu'p',\nu p } d_{\nu p } , \label{eq : phat}\ ] ] where the repeated symbols are summed . here , @xmath785 is the measured temperature at a pixel @xmath786 in a frequency band @xmath746 , @xmath787 is a map of an annulus corresponding to a given angular bin @xmath782 , which has been convolved with the beam and scaled by the frequency dependence of the sz effect , @xmath788 is the noise covariance matrix ( which is taken to be diagonal in pixel space and @xmath746 , i.e. , @xmath789 ) , and @xmath790 is the signal covariance matrix of cmb convolved with the beam ( @xmath494 and @xmath791 are the cmb power spectrum and the beam transfer function , respectively ) .
a matrix @xmath792 gives the @xmath793 error of @xmath783 as @xmath794 , and is given by @xmath795^{-1}_{\nu'p',\nu p } ( t_\beta)_{\nu p}.\ ] ] for @xmath785 , we use the foreground - cleaned v- and w - band temperature maps at the healpix resolution of @xmath796 , masked by the _ kq75y7 _ mask . note
that the _ kq75y7 _ mask includes the 7-year source mask , which removes a potential bias in the reconstructed profile due to any sources which are bright enough to be resolved by _ wmap _ , as well as the sources found by other surveys .
specifically , the 7-year point source mask includes sources in the 7-year _ wmap _ source catalog @xcite ; sources from @xcite ; sources with 22 ghz fluxes @xmath797 jy from @xcite ; flat spectrum objects from @xcite ; and sources from the blazar survey of @xcite and @xcite . in figure [
fig : coma ] , we show the measured angular radial profiles of coma in 16 angular bins ( separated by @xmath798 arcmin ) , in units of the rayleigh - jeans temperature , for the v- and w - band data , as well as for the v+w combined data . the error bar at a given angular bin is given by @xmath794 .
we find that all of these measurements agree well at @xmath799 arcmin ; however , at smaller angular scales , @xmath800 arcmin , the v+w result shows _ less _ sz than both the v- and w - only results .
does this make sense ?
as described in appendix [ app : sz ] , our optimal estimator uses both the @xmath720-weighted v+w map and the @xmath801-weighted v@xmath231w map .
while the latter map vanishes for cmb , it does not vanish for the sz effect .
therefore , the latter map can be used to separate cmb and sz effectively .
this explains why the v+w result and the other results are different only at small angular scales : at @xmath799 arcmin , the measured signal is @xmath802k .
if this was due to sz , the difference map , v@xmath231w , would give @xmath803k , which is smaller than the noise level in the difference map , and thus would not show up .
in other words , our estimator can not distinguish between cmb and sz at @xmath804 arcmin .
on the other hand , at @xmath805 arcmin , each of the v- and w - band data shows much bigger signals , @xmath806k .
if this was due to sz , the difference map would give @xmath807k , which is comparable to or greater than the noise level in the difference map , and thus would be visible .
we find that the difference map does not detect signals in @xmath808 arcmin , which suggests that the measured signal , @xmath809k , is _ not _ due to sz , but due to cmb . as a result ,
the v@xmath190w result shows less sz than the v- and w - only results . in order to quantify a statistical significance of detection and interpret the result , we model the sz profile using a spherical @xmath769 model @xcite : @xmath810^{(1 - 3\beta)/2}.\ ] ] to make our analysis consistent with previous measurements described later , we fix the core radius , @xmath811 , and the slope parameter , @xmath769 , at @xmath770 arcmin and @xmath771 @xcite , and vary only the central decrement , @xmath812 . in this case , the optimal estimator is @xmath813^{-1}_{\nu'p',\nu p } d_{\nu p},\ ] ] where @xmath814 is a map of the above @xmath769 model with @xmath815 , and @xmath816^{-1}_{\nu'p',\nu p}t_{\nu p } , \label{eq : f}\ ] ] gives the @xmath793 error as @xmath817 . for v+w
, we find @xmath818 which is a @xmath741 measurement of the sz effect toward coma . in terms of the compton @xmath371-parameter at the center
, we find @xmath819 let us compare this measurement with the previous measurements .
@xcite used the 5.5-m telescope at the owens valley radio observatory ( ovro ) to observe coma at 32 ghz . using the same @xmath811 and @xmath769 as above
, they found the central decrement of @xmath820k ( 68% cl ) , after subtracting @xmath821k due to point sources ( 5c4.81 and 5c4.85 ) .
these sources have been masked by our point - source mask , and thus we do not need to correct for point sources . while our estimate of @xmath822 is different from that of @xcite only by @xmath823 , and thus is statistically consistent , we note that @xcite did not correct for the cmb fluctuation in the direction of coma . as the above results indicate that the cmb fluctuation in the direction of coma is on the order of @xmath809k , it is plausible that the ovro measurement implies @xmath824k , which is an excellent agreement with the _ wmap _ measurement . the coma cluster has been observed also by the millimetre and infrared testagrigia observatory ( mito ) experiment @xcite . using the same @xmath811 and @xmath769 as above , @xcite found @xmath825 , @xmath826 , and @xmath827k ( 68% cl ) at 143 , 214 , and 272 ghz , respectively , in units of thermodynamic temperatures .
as mito has 3 frequencies , they were able to separate sz , cmb , and the atmospheric fluctuation . by fitting these 3 data points to the sz spectrum , @xmath828
, we find @xmath829 , which is an excellent agreement with the _ wmap _ measurement .
the first error is statistical and the second error is systematic due to 10% calibration error of mito .
the calibration error of the _ wmap _ data ( 0.2% ; * ? ? ?
* ) is negligible .
finally , one may try to fit the multi - wavelength data of @xmath830 to separate the sz effect and cmb . for this purpose
, we fit the _ wmap_data in v- and w - band to the @xmath769 model _ without correcting for the cmb fluctuation_. we find @xmath831k and @xmath832k in thermodynamic units ( 68% cl ) .
the ovro measurement , @xmath833k @xcite , has been scaled to the rayleigh - jeans temperature with the sz spectral dependence correction , and thus we use this measurement at @xmath834 . fitting the _ wmap _ and ovro data to the sz effect plus cmb , and the mito data only to the sz effect ( because the cmb was already removed from mito using their multi - band data ) , we find @xmath835 and @xmath836k ( 68% cl ) .
this result is consistent with our interpretation that the @xmath371-parameter of the center of coma is @xmath837 and the cmb fluctuation is on the order of @xmath809k .
the analysis presented here shows that our optimal estimator is an excellent tool for extracting the sz effect from multi - frequency data .
the coma cluster is the brightest sz cluster on the sky .
there are other clusters that are bright enough to be seen by _
wmap_. in order to select candidates , we use the sample of 49 nearby clusters compiled by @xcite , which are used by the cosmological analysis given in @xcite .
these clusters are selected from the _ rosat _ all - sky survey , and have detailed follow - up observations by _ chandra_. the latter property is especially important , as it allows us to _ directly _ compare the measured sz effect in the _ wmap _ data and the expected one from the x - ray data on a cluster - by - cluster basis , without relying on any scaling relations . , which uses a larger number of clusters but relies on scaling relations . nevertheless , the results obtained from the analysis in this section and those in section [ sec : szstacking ] are in good agreement . ]
not all nearby clusters in @xcite are suitable for our purpose , as some clusters are too small to be resolved by the _
we thus select the clusters that have the radius greater than @xmath838 on the sky : specifically , we use the clusters whose @xmath839 is greater than @xmath838 . here , @xmath774 is the radius within which the mean overdensity is 500 times the critical density of the universe .
we find that 38 clusters satisfy this condition .
( note that the coma cluster is not included in this sample . ) of these , 5 clusters have @xmath840 , 7 clusters have @xmath841 , 13 clusters have @xmath842 , and 13 clusters have @xmath843 . here
, @xmath844 is the mass enclosed within @xmath774 , i.e. , @xmath845 .
finally , we remove the clusters that lie within the _ kq75y7 _ mask ( including the diffuse and the source mask ) , leaving 29 clusters for our analysis .
( 1 cluster ( a478 ) in @xmath841 , 4 clusters in @xmath842 , and 4 clusters in @xmath843 are masked , mostly by the point source mask . )
the highest redshift of this sample is @xmath846 ( a2142 ) .
lccc @xmath847 & 5 & @xmath848 & @xmath849@xmath850 & 6 & @xmath851 & @xmath852 @xmath853 & 9 & @xmath854 & @xmath855 @xmath856 & 9 & @xmath857 & @xmath858 @xmath859 & 11 & @xmath860 & @xmath861@xmath862 & 18 & @xmath863 & @xmath864 @xmath859 & & & cooling flow & 5 & @xmath865 & @xmath866non - cooling flow & 6 & @xmath867 & @xmath868@xmath869 & 20 & @xmath870 & @xmath871@xmath872 & 29 & @xmath873 & @xmath874 [ tab : vikh_vs_arnaud ] in figure [ fig : massive ] , we show the measured sz effect in the symbols with error bars , as well as the expected sz from the x - ray data in the dashed lines . to compute the expected sz , we use equation ( [ eq : sz ] ) with @xmath875 , where @xmath577 and @xmath876 are fits to the x - ray data .
specifically , we use ( see equation ( 3 ) and ( 8) of * ? ? ?
* ) @xmath877^{0.45}},\end{aligned}\ ] ] where @xmath878 . the parameters in the above equations are found from the _ chandra _ x - ray data , and kindly made available to us by a. vikhlinin . for a given pressure profile , @xmath749
, we compute the sz temperature profile as @xmath879 , \label{eq : tszfit}\end{aligned}\ ] ] where @xmath880 is the projected electron pressure profile on the sky : @xmath881 here , we truncate the pressure profile at @xmath882 . we take this to be @xmath883 . while the choice of the boundary is somewhat arbitrary ,
the results are not sensitive to the exact value because the pressure profile declines fast enough .
we find a good agreement between the measured and expected sz signals ( see figure [ fig : massive ] ) , except for a754 : a754 is a merging cluster with a highly disturbed x - ray morphology , and thus the expected sz profile , which is derived assuming spherical symmetry ( equation [ eq : tszfit ] ) , may be different from the observed one . to make the comparison quantitative
, we select clusters within a given mass bin , and fit the expected sz profiles to the _ wmap _ data with a single amplitude , @xmath884 , treated as a free parameter .
the optimal estimator for the normalization of pressure , @xmath884 , is @xmath885^{-1}_{\nu'p',\nu p } d_{\nu p } , \label{eq : estimate_a}\ ] ] where @xmath814 is a map containing the predicted sz profiles around clusters , and the @xmath793 error is @xmath817 where @xmath886 is given by equation ( [ eq : f ] ) .
we summarize the results in the second column of table [ tab : vikh_vs_arnaud ] .
we find that the amplitudes of _ all _ mass bins are consistent with unity ( @xmath887 ) to within @xmath324 ( except for the `` non - cooling flow '' case , for which @xmath884 is less than unity at @xmath888 ; we shall come back to this important point in the next section ) .
the agreement is especially good for the highest mass bin ( @xmath840 ) , @xmath889 ( 68% cl ) .
note that this is a @xmath890 detection of the sz effect , just from stacking 5 clusters . by stacking 11 clusters with @xmath891 ( i.e. , all clusters in figure [ fig : massive ] but coma ) , we find @xmath892 ( 68% cl ) , a @xmath743 detection .
in other words , one does not need to stack many tens or hundreds of clusters to see the sz effect in the _ wmap _ data , contrary to what is commonly done in the literature @xcite . from this study
, we conclude that the _ wmap _ data and the expectation from the x - ray data are in good agreement . recently , @xcite derived pressure profiles of 33 clusters from the x - ray follow - up observations of the rexcess clusters using _ xmm - newton_. the rexcess sample contains clusters selected from the _ rosat _ all - sky survey @xcite . by scaling the pressure profiles appropriately by mass and redshift and taking the median of the scaled profiles
, they produced a `` universal pressure profile . ''
we describe this profile in appendix [ sec : arnaud ] .
we show the predicted @xmath893 from arnaud et al.s pressure profile in figure [ fig : massive ] ( solid lines ) . in order to compute their profile ,
we need the mass of clusters , @xmath844 .
we take @xmath844 from the the 6th column of table 2 in @xcite , which are derived from the so - called mass-@xmath894 relation , the most precise mass proxy known to date with a scatter of about 5% . , with @xmath895 for coma s redshift , @xmath896 .
we use the x - ray temperature of @xmath897 kev @xcite . ]
again , we take the outer boundary of the pressure to be @xmath883 .
we fit arnaud et al s profiles to the _ wmap_data of 29 clusters .
we find that , in all but one of the mass bins , the best - fitting normalization , @xmath884 , is less than unity by more than @xmath324 . by stacking 11 clusters with @xmath898 , we find @xmath899 ( 68% cl ) .
this measurement rules out @xmath887 by @xmath900 .
the universal pressure profile overestimates the sz effect by @xmath901% . what causes the discrepancy ?
the thermal / dynamical state of gas in clusters may be the culprit . from figure
[ fig : massive ] , we find that the x - ray data ( hence the sz effect ) and the universal profile agree well for `` cooling flow '' clusters , but do not agree for non - cooling flow clusters .
the cooling flow clusters have _ cool cores _ ,
in which the cooling time ( due to bremsstrahlung ) is shorter than the hubble time @xcite .
the clusters shown in figure [ fig : massive ] are classified as either `` cooling flow '' or `` non - cooling flow '' clusters , following the definition of @xcite .
we find that arnaud et al.s profiles agree with the x - ray data on the individual clusters well at @xmath902 .
this agrees with figure 8 of @xcite .
the profiles differ significantly in the inner parts of clusters , which is also in good agreement with the conclusion of @xcite : they find that cool - core clusters show much steeper inner profiles than non - cool - core clusters ( their figure 2 and 5 ) . for cooling - flow clusters ,
the agreement between the _ wmap _ data and arnaud et al.s profile is good : @xmath903 ( 68% cl ) .
however , for non - cooling - flow clusters , we find a very low amplitude , @xmath904 ( 68% cl ) , which rules out arnaud et al.s profile by @xmath905 .
a similar trend is also observed for the individual x - ray data of vikhlinin et al .
: @xmath906 and @xmath867 ( 68% cl ) for cooling - flow and non - cooling - flow clusters , respectively ; however , statistical significance is not large enough to exclude @xmath887 .
based on this study , we conclude that one must distinguish between cool - core ( cooling flow ) and non - cool - core clusters when interpreting the observed profile of the sz effect .
it is clear ( at the @xmath900 level ) that arnaud et al.s profile is inconsistent with the individual x - ray data and the sz data taken by _
wmap _ , and ( at the @xmath905 level ) _
one must distinguish between the cool - core and non - cool - core clusters_. interestingly , _ all _ of cooling - flow clusters are `` relaxed '' clusters , and _ all _ of non - cooling - flow clusters are `` non - relaxed '' ( i.e. , morphologically disturbed ) clusters , according to the criterion of @xcite . if we interpret this as non - cooling - flow clusters having undergone recent mergers , then we may conclude that we are finding the effect of mergers on the sz effect .
while our conclusion is still based on a limited number of clusters , it may be valid for a much larger sample of clusters , as we shall show in section [ sec : arnaudresults ] .
finally , we note that the current generation of hydrodynamical simulations predict the pressure profiles that are even steeper than arnard et al.s profile ( see figure 7 of * ? ? ?
therefore , the simulations also overpredict the amount of pressure in clusters relative to the _ wmap _ data .
we shall come back to this point in section [ sec : ksandsim ] .
k ) , at @xmath907 , 35 , 63 , and 91 arcmin . the v- and w - band data are combined using the optimal estimator .
( top ) the sz effect measured from the locations of clusters of galaxies .
the results with three different maximum redshifts , @xmath908 ( blue ; left ) , 0.2 ( green ; middle ) , and 0.45 ( red ; right ) , are shown .
the error bars include noise due to the cmb fluctuation , and thus are correlated ( see equation ( [ eq : corr ] ) for the correlation matrix ) .
( middle ) a null test showing profiles measured from random locations on the sky ( for @xmath909 ; the number of random locations is the same as the number of clusters used in the top panel ) .
three random realizations are shown .
our method does not produce biased results .
( bottom ) the measured profile ( @xmath909 ) is compared with the model profiles derived from the median of 33 clusters in the rexcess sample @xcite and theoretically calculated from hydrostatic equilibrium @xcite with two different concentration parameters . note that the model profiles are calculated also for @xmath909 but have not been multiplied by the best - fitting normalization factors given in table [ tab : szamplitude ] .
the theoretical profiles are processed in the same manner that the data are processed , using equation ( [ eq : phat ] ) .
, width=287 ] to explore the sz effect in a large number of clusters , we use a galaxy cluster catalog consisting of the _ rosat_-eso flux - limited x - ray ( reflex ) galaxy cluster survey @xcite in the southern hemisphere above the galactic plane ( @xmath910 and @xmath911 ) and the extended brightest cluster sample ( ebcs ; * ? ? ?
* ; * ? ? ?
* ) in the northern hemisphere above the galactic plane ( @xmath912 and @xmath911 ) .
some clusters are contained in both samples . eliminating the overlap
, this catalog contains 742 clusters of galaxies . of these , 400 , 228 , and
114 clusters lie in the redshift ranges of @xmath913 , @xmath914 , and @xmath915 , respectively .
we use the foreground - reduced v- and w - band maps at the healpix resolution of @xmath796 , masked by the _ kq75y7 _ mask , which eliminates the entire virgo cluster . note
that this mask also includes the point - source mask , which masks sources at the locations of some clusters ( such as coma ) . after applying the mask , we have 361 , 214 , and 109 clusters in @xmath913 , @xmath914 , and @xmath915 , respectively .
we again use equation ( [ eq : phat ] ) to find the angular radial profile in four angular bins . for this analysis
, @xmath787 is a map containing many annuli ( one annulus around each cluster ) corresponding to a given angular bin @xmath782 , convolved with the beam and scaled by the frequency dependence of the sz effect .
we show the measured profile in the top panel of figure [ fig : szstacking ] . we have done this analysis using 3 different choices of the maximum redshift , @xmath916 , to select clusters : @xmath908 , 0.2 , and 0.45 .
we find that the results are not sensitive to @xmath916 .
as expected , the results for @xmath917 have the largest error bars .
the error bars for @xmath918 and 0.45 are similar , indicating that we do not gain much more information from @xmath919 .
the error bars have contributions from instrumental noise and cmb fluctuations .
the latter contribution correlates the errors at different angular bins .
the top panel shows a decrement of @xmath920k at a very large angular distance from the center , @xmath921 arcmin , for @xmath918 . as we do not expect to have such an extended gas distribution around clusters ,
one may wonder if this result implies that we have a bias in the zero level . in order to check for a potential systematic bias , we perform the following null test : instead of measuring the sz signals from the locations of clusters , we measure them from random locations in the _ wmap _ data . in the middle panel of figure [ fig : szstacking ] , we show that our method passes a null test .
we find that the measured profiles are consistent with zero ; thus , our method does not introduce a bias .
is this signal at a degree scale real ?
for example , are there nearby massive clusters ( such as coma ) which give a significant sz effect at a degree scale ? while the virgo cluster has the largest angular size on the sky , the _ kq75y7 _ mask eliminates virgo . in order to see
if other nearby clusters give significant contributions , we remove all clusters at @xmath922 ( where there are 57 clusters ) and remeasure the sz profile .
we find that the changes are small , less than @xmath923k at all angular bins . at @xmath924arcmin ,
the change is especially small , @xmath925k , and thus nearby clusters do not make much contribution to this bin .
the apparent decrement at @xmath921 arcmin is probably due to a statistical fluctuation .
the angular bins are correlated with the following correlation matrix : @xmath926 where the columns correspond to @xmath907 , 35 , 63 , and 91 arcmin , respectively .
the decrements at the first two bins ( at @xmath907 and 35 arcmin ) can drive the third bin at @xmath921 arcmin to be more negative .
note also that one of the realizations shown in the bottom panel ( `` random 1 '' in the middle panel of figure [ fig : szstacking ] ) shows @xmath927k at @xmath921 arcmin .
the second bin is also negative with a similar amplitude . on the other hand , ``
random 2 '' shows both positive temperatures at the second and third bins , which is also consistent with a positive correlation between these bins .
finally , in the bottom panel of figure [ fig : szstacking ] , we compare the measured sz profile with the expected profiles from various cluster gas models ( described in section [ sec : szimplications ] ) . none of them show a significant signal at @xmath921 arcmin , which is also consistent with our interpretation that it is a statistical fluctuation . in order to interpret the measured sz profile ,
we need a model for the electron pressure profile , @xmath749 ( see equation ( [ eq : sz ] ) ) . for fully ionized gas
, the electron pressure is related to the _ gas _ ( baryonic ) pressure , @xmath928 , by @xmath929 where @xmath930 is the abundance of hydrogen in clusters . for @xmath931 ,
one finds @xmath932 .
we explore three possibilities : ( i ) arnaud et al.s profile that we have used in section [ sec : nearby ] , ( ii ) theoretical profiles derived by assuming that the gas pressure is in hydrostatic equilibrium with gravitational potential given by an navarro - frenk - white mass density profile @xcite , and ( iii ) theoretical profiles from hydrodynamical simulations of clusters of galaxies with and without gas cooling and star formation @xcite .
the case ( ii ) is relevant because this profile is used in the calculation of the sz power spectrum @xcite that has been used as a template to marginalize over in the cosmological parameter estimation since the 3-year analysis @xcite .
analytical models and hydrodynamical simulations for the sz signal are also the basis for planned efforts to use the sz signal to constrain cosmological models .
as we have shown in the previous section using 29 nearby clusters , arnaud et al.s pressure profile overpredicts the sz effect in the _ wmap _ data by @xmath901% .
an interesting question is whether this trend extends to a larger number of clusters . .
the thick horizontal line shows the cosmic mean baryon fraction , @xmath933 .
the solid line shows the gas mass fraction , @xmath934 , derived from x - ray observations @xcite , while the dotted and dashed lines show @xmath935 predicted from hydrostatic equilibrium @xcite with nfw concentration parameters of @xcite and @xcite , respectively .
, width=302 ] the normalization of the ks profile has been fixed by assuming that the gas density at the virial radius is equal to the cosmic mean baryon fraction , @xmath936 , times the total mass density at the virial radius .
this is an upper limit : for example , star formation turns gas into stars , reducing the amount of gas .
ks also assumes that the gas is virialized and in thermal equilibrium ( i.e. , electrons and protons share the same temperature ) everywhere in a cluster , that virialization converts potential energy of the cluster into thermal energy only , and that the pressure contributed by bulk flows , cosmic rays , and magnetic fields are unimportant .
we give details of the gas pressure profiles in appendix [ sec : pressure ] . in the top left panel of figure [ fig : profile ] , we show arnaud et al.s pressure profiles ( see appendix [ sec : arnaud ] ) in the solid lines , and the ks profiles ( see appendix [ sec : ks ] ) in the dotted and dashed lines .
one of the inputs for the ks profile is the so - called concentration parameter of the nfw profile .
the dotted line is for the concentration parameter of @xmath937 @xcite , which was used by @xcite for their calculation of the sz power spectrum . here
, @xmath938 is the virial mass , i.e. , mass enclosed within the virial radius .
the dashed line is for @xmath939 , which was found from recent n - body simulations with the _ wmap _
5-year cosmological parameters @xcite .
we find that the ks profiles and arnaud et al.s profiles generally agree .
the agreement is quite good especially for the ks profile with the concentration parameter of @xcite .
the ks profiles tend to overestimate the gas pressure relative to arnaud et al.s one for low - mass clusters ( @xmath940 ) .
can we explain this trend by a smaller gas mass fraction in clusters than the cosmic mean ? to answer this , we compute the gas mass fraction by integrating the gas density profile : @xmath941 where @xmath844 and @xmath942 are the total mass and gas mass contained within @xmath774 , respectively . in figure [
fig : fgas ] , we show @xmath935 from x - ray observations @xcite : @xmath943 for @xmath944 , and @xmath935 from the ks profiles with the concentration parameters of @xcite and @xcite .
we find that the ks predictions , @xmath945 , are always much smaller than the cosmic mean baryon fraction , @xmath933 , and are nearly independent of mass .
a slight dependence on mass is due to the dependence of the concentration parameters on mass . while the ks profile is normalized such that the gas density _ at _ the virial radius is @xmath936 times the total mass density , the gas mass _ within _ @xmath774 is much smaller than @xmath936 times @xmath844 , as the gas density and total matter density profiles are very different near the center : while the gas density profile has a constant - density core , the total matter density , which is dominated by dark matter , increases as @xmath946 near the center .
however , the behavior of @xmath935 measured from x - ray observations is very different .
it has a much steeper dependence on mass than predicted by ks .
the reason for such a steep dependence on mass is not yet understood .
it could be due to star formation occurring more effectively in lower mass clusters . in any case
, for @xmath947 , the observed gas mass fraction is @xmath948 , which is only 10% smaller than the ks value , @xmath949 . for @xmath950 , the observed gas mass fraction , @xmath951 ,
is about 30% smaller than the ks value .
this is consistent with the difference between the ks and arnaud et al.s pressure profiles that we see in figure [ fig : profile ] ; thus , once the observed mass dependence of @xmath935 is taken into account , these profiles agree well .
to calibrate the amplitude of gas pressure , we shall use the ks pressure profile ( without any modification to @xmath935 ) as a template , and find its normalization , @xmath884 , from the _ wmap _ data using the estimator given in equation ( [ eq : estimate_a ] ) .
we shall present the results for hydrodynamical simulations later . for a given gas pressure profile , @xmath928 , we compute the electron pressure as @xmath952 ( see equation ( [ eq : pgaspe ] ) ) .
we then use equation ( [ eq : tszfit ] ) to calculate the expected sz profile , @xmath893 .
we take the outer boundary of the pressure to be 3 times the virial radius , @xmath953 , which is the same as the parameter used by @xcite . in the right panels of figure
[ fig : profile ] , we show the predicted @xmath893 , which will be used as templates , i.e. , @xmath814 . estimated from clusters in the catalog using the measured x - ray luminosities in 0.1 - 2.4 kev band , @xmath954 , and equations ( [ eq : r500 ] ) and ( [ eq : m500 ] ) . the light blue , dark blue , and pink histograms show @xmath955 , 0.2 , and 0.1 , respectively .
, width=321 ] now , in order to compute the expected pressure profiles from each cluster in the catalog , we need to know @xmath774 .
we calculate @xmath774 from the observed x - ray luminosity in _ rosat _ s 0.1 - 2.4 kev band , @xmath954 , as @xmath956 where @xmath957^{1/2}$ ] for a @xmath1cdm model .
this is an empirical relation found from x - ray observations ( see equation ( 2 ) of * ? ? ?
* ) based upon the temperature-@xmath954 relation from @xcite and the @xmath774-temperature relation from @xcite .
the error bars have been calculated by propagating the errors in the temperature-@xmath954 and @xmath774-temperature relations .
admittedly , there is a significant scatter around this relation , which is the most dominant source of systematic error in this type of analysis .
( the results presented in section [ sec : nearby ] do not suffer from this systematic error , as they do not rely on @xmath954-@xmath774 relations . ) as @xmath958 , a @xmath959% error in the predicted values of @xmath774 gives the mass calibration error of @xmath960% .
moreover , the sz effect is given by @xmath844 times the gas temperature , the latter being proportional to @xmath961 according to the virial theorem .
therefore , the total calibration error can be as big as @xmath962% . in order to quantify this systematic error , we repeat our analysis for 3 different size - luminosity relations : ( i ) the central values , ( ii ) the normalization and slope shifted up by @xmath793 to @xmath963 and @xmath964 , and ( iii ) the normalization and slope shifted down by @xmath793 to @xmath965 and @xmath966 .
we adopt this as an estimate for the systematic error in our results due to the size - luminosity calibration error . for how this error would affect our conclusions ,
see section [ sec : sz_syserror ] . note that this estimate of the systematic error is conservative , as we allowed _ all _ clusters to deviate from the best - fit scaling relation at once by @xmath967 .
in reality , the nature of this error is random , and thus the actual error caused by the scatter in the scaling relation would probably be smaller .
@xcite performed such an analysis , and found that the systematic error is sub - dominant compared to the statistical error .
nevertheless , we shall adopt our conservative estimate of the systematic error , as the mean scaling relation also varies from authors to authors .
the mean scaling relations used by @xcite are within the error bar of the scaling relation that we use ( equation ( [ eq : r500 ] ) ) . in figure
[ fig : counts ] , we show the distribution of @xmath844 estimated from clusters in the catalog using the measured values of @xmath954 and equations ( [ eq : r500 ] ) and ( [ eq : m500 ] ) . the distribution peaks at @xmath968 for @xmath909 and 0.45 , while it peaks at @xmath969 for @xmath908 .
lccccc @xcite & x - ray obs .
& @xmath970 & @xmath971 & @xmath972 & @xmath973 @xcite & rexcess scaling & n / a & @xmath974 & @xmath975 & @xmath976 @xcite & intrinsic scaling & n / a & @xmath977 & @xmath978 & @xmath979 @xcite & @xmath980 & n / a & @xmath981 & @xmath972 & @xmath973 @xcite & @xmath982 & n / a & @xmath983 & @xmath984 & @xmath985 @xcite & equation ( [ eq : c_seljak ] ) & @xmath986 & @xmath987 & @xmath988 & @xmath989 @xcite & equation ( [ eq : c_duffy ] ) & @xmath972 & @xmath990 & @xmath991 & @xmath973 @xcite & non - radiative & n / a & @xmath992 & @xmath993 & @xmath994 @xcite & cooling+sf & n / a & @xmath995 & @xmath996 & @xmath997 [ tab : szamplitude ] for arnaud et al.s pressure profile , we find the best - fitting amplitudes of @xmath998 and @xmath981 ( 68% cl ) for @xmath917 and 0.2 , respectively .
the former result is fully consistent with what we find from the nearby clusters in section [ sec : nearby ] : @xmath999 ( 68% cl ; for @xmath1000 and @xmath742 ) .
the significance level of statistical detection of the sz effect is about 8@xmath9 for @xmath909 . with
the systematic error included , we find @xmath1001 for @xmath909 ; however , the above agreement may suggest that the fiducial scaling relation ( equation ( [ eq : r500 ] ) ) is , in fact , a good one . as we have shown in section [ sec : nearby ] , the measured sz effects and the predictions from the x - ray data agree on a cluster - by - cluster basis .
a plausible explanation for the discrepancy between the _ wmap _ data and arnaud et al.s profile is that arnaud et al.s profile does not distinguish between cooling - flow and non - cooling - flow clusters . nevertheless , this result , which shows that the sz effect seen in the _ wmap _ data is _ less _ than the average `` expectation '' from x - ray observations , agrees qualitatively with some of the previous work @xcite .
the other work showed that the sz effect seen in the _ wmap _ data is consistent with expectations from x - ray observations @xcite .
these authors used widely different methods and cluster catalogues .
@xcite were the first to claim that the sz effect seen in the _ wmap _ data is significantly less than expected from x - ray data , by using 31 clusters compiled by @xcite .
@xcite extended the analysis of @xcite by using 38 clusters compiled by @xcite , for which the observational data of the sz effect from ovro and berkeley illinois maryland association ( bima ) are available .
they did not use scaling relations , but used a spherical isothermal @xmath769 model to fit the x - ray surface brightness profile of each cluster in the catalog , and calculated the expected sz signals , assuming that the intracluster gas is isothermal .
@xcite found that the measured signal is smaller than expected from x - ray data by a factor of 34 , and @xcite found a similar result for the cluster catalog of @xcite .
@xcite used the same cluster catalog that we use ( reflex+ebcs ) , but used a different scaling relation : they related the cluster core radius to the x - ray luminosity ( we relate @xmath774 to the x - ray luminosity ) .
they found a large discrepancy ( similar to * ? ? ?
* ; * ? ? ?
* ) when a spherical isothermal @xmath769 model was used to predict the sz signal , while they found a smaller discrepancy ( similar to our results ) when more realistic gas models were used .
@xcite used 193 clusters selected from the xbac catalog .
their catalog consisted of the clusters that have measured x - ray temperatures ( @xmath1002 kev ) .
they then used a scaling relation between @xmath1003 and the x - ray temperature .
they found that the measured sz signal and x - ray data are consistent .
@xcite used the 5-year _ wmap _ data and a bigger sample of 893 clusters and a scaling relation between @xmath774 and the x - ray luminosity taken from @xcite .
they compared the measured integrated pressure from the _ wmap _ data to the expectation from arnaud et al.s profile , and concluded that they agree very well .
( the normalization is consistent with unity within the statistical uncertainty . )
we find , on the other hand , that the normalization is significantly less than unity compared to the statistical uncertainty .
how can we reconcile these results ?
one possibility would be the difference in the scaling relations .
the scaling relation shifted down by @xmath793 would make the predicted sz signals smaller , which would then increase the best - fitting amplitude .
given the size of the systematic error , @xmath1004 , @xmath1005 may not be inconsistent with the data . specifically , they used two scaling relations : * @xmath1006^{0.222}$ ] , * @xmath1007^{0.207}$ ] , where the relations 1 and 2 correspond to the `` rexcess '' and `` intrinsic '' relations in @xcite , respectively . here , @xmath1008 is the x - ray luminosity measured within @xmath774 , which is calculated from @xmath954 . while we do not have the conversion factors they used ,
a typical magnitude of the conversion factors is about 10% , according to @xcite . a 10% change in @xmath954 gives a 2% change in @xmath774 , which is negligible compared to the other uncertainties ; thus , we shall assume that @xmath954 and @xmath1008 are the same , and repeat our analysis using these scaling relations .
we find the amplitudes of @xmath1009 and @xmath977 ( @xmath909 ; 68% cl ) for the relations 1 and 2 , respectively ; thus , while these scaling relations give larger amplitudes , they can not completely explain the difference between the results of @xcite ( @xmath1010 ) and our results . however , we find that the discrepancy is much less for high x - ray luminosity clusters .
see section [ sec : lum_bin ] .
while the method of @xcite and our method are similar , they are different in details .
we compare the predicted angular radial profiles of the sz effect to the _ wmap _ data to find the best - fitting amplitude .
@xcite measured the _ integrated _ pressure within 5 times @xmath774 , and converted it to the integrated pressure within @xmath774 , @xmath1011 , assuming the distribution of pressure beyond @xmath774 is described by the profile of @xcite .
whether the difference in methodology can account for the difference between our results and their results is unclear , and requires further investigation.[multiblock footnote omitted ] in any case , we emphasize once again that the sz effect measured by the _ wmap _ and the predictions from x - ray data agree well , _ when the actual x - ray profile of individual clusters , rather than the average ( or median ) profile , is used _ , and there is a reason why arnaud et al.s profile would overpredict the pressure ( i.e. , cooling flows ; see section [ sec : nearby ] ) .
therefore , it is likely that the difference between our results and @xcite simply points to the fundamental limitation of the analysis using many clusters ( with little or no x - ray data ) and scaling relations .
let us turn our attention to the analytical ks profile .
for the ks profile with the concentration parameter of @xcite , we find the best - fitting amplitudes of @xmath1012 and @xmath987 ( 68% cl ) for @xmath908 and 0.2 , respectively .
for the ks profile with the concentration parameter of @xcite , we find @xmath1013 and @xmath990 ( 68% cl ) for @xmath908 and 0.2 , respectively .
these results are consistent with those for arnaud et al.s pressure profiles .
recently , the spt collaboration detected the sz power spectrum at @xmath18 . by fitting their sz power spectrum data to the theoretical model of @xcite
, they found the best - fitting amplitude of @xmath277 ( 68% cl ; * ? ? ? * ) .
the calculation of @xcite is based on the ks gas pressure profile .
as the amplitude of sz power spectrum is proportional to the gas pressure squared , i.e. , @xmath1014 , our result for the ks profiles , @xmath1015 , is consistent with @xmath1016 found from spt .
the act collaboration placed an upper limit of @xmath1017 ( 95% cl ; * ? ? ? * ) , which is consistent with the spt result .
what do hydrodynamical simulations tell us ? as the analytical calculations such as @xcite are limited , we also fit the pressure profiles derived from hydrodynamical simulations of @xcite to the _ wmap _ data . in the bottom panels of figure [ fig : profile ] ,
we show the gas pressure profiles from `` non - radiative '' and `` cooling+star formation ( sf ) '' runs . by fitting the sz templates constructed from these simulated profiles to the _ wmap _ data
, we find the best - fitting amplitudes of @xmath992 and @xmath1018 ( 68% cl ) for non - radiative and cooling+sf runs , respectively , which are consistent with the amplitudes found for the ks profiles and arnaud et al.s profiles .
see table [ tab : szamplitude ] for a summary of the best - fitting amplitudes .
that the ks , simulation , and arnaud et al.s profiles yield similar results indicates that _ all _ of these profiles overpredict the amount of sz effect seen in the _ wmap _ data by @xmath1019% .
this conclusion is made robust by the results we presented in section [ sec : nearby ] : the analysis that does _ not _ use scaling relations between @xmath954 and @xmath774 , but uses only a subset of clusters that have the detailed follow - up observations by _ chandra _ , yields the same result .
this is one of the main results of our sz analysis . to see the dependence of the best - fitting normalization on x - ray luminosities ( hence @xmath844 ) , we divide the cluster samples into 3 luminosity bins : ( i ) `` high @xmath954 '' with @xmath1020 , ( ii ) `` low @xmath954 '' with @xmath1021 , and ( iii ) clusters fainter than ( ii )
there are 82 , 417 , and 129 clusters in ( i ) , ( ii ) , and ( iii ) , respectively . in table
[ tab : szamplitude ] , we show that we detect significant sz signals in ( i ) and ( ii ) , despite the smaller number of clusters used in each luminosity bin
. we do not have a statistically significant detection in ( iii ) .
the high @xmath954 clusters have @xmath1022 . for these clusters ,
the agreement between the _ wmap _ data and the expected sz signals is much better . in particular , for the rexcess scaling relation , we find @xmath1023 , which is consistent with unity within the @xmath793 statistical error .
this implies that , at least for high x - ray luminosity clusters , our results and the results of @xcite agree within the statistical uncertainty . on the other hand , we find that less luminous clusters tend to have significantly lower best - fitting amplitudes for all models of gas - pressure profiles and scaling relations that we have explored .
this trend is consistent with , for example , the gas mass fraction being lower for lower mass clusters .
it is also consistent with radio point sources filling some of the sz effect seen in the _ wmap _ data .
for the point source contamination , see section [ sec : sz_syserror ] .
the best - fitting amplitudes may be shifted up and down by @xmath962% due to the calibration error in the size - luminosity relation ( equation ( [ eq : r500 ] ) ) . as we have shown already ,
the best - fitting amplitudes for the ks profiles can be shifted up to @xmath1024 and @xmath1025 for the concentration parameters of @xcite and @xcite , respectively .
similarly , the amplitude for arnaud et al.s profile can be shifted up to @xmath1026 .
as this calibration error shifts all amplitudes given in table [ tab : szamplitude ] by the same amount , it does not affect our conclusion that all of the gas pressure profiles considered above yield similar results .
this type of systematic error can be reduced by using a subset of clusters of galaxies for which the scaling relations are more tightly constrained ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) ; however , reducing the number of samples increases the statistical error .
indeed , the analysis presented in section [ sec : nearby ] does not suffer from the ambiguity in the scaling relations
. how important are radio point sources ?
while we have not attempted to correct for potential contamination from unresolved radio point sources , we estimate the magnitude of effects here . if , on average , each cluster has a @xmath1027 mjy source , then the corresponding temperatures , @xmath1028,\ ] ] are 2.24 , 2.29 , and 2.19 @xmath1029k in q , v , and w bands , respectively . here ,
@xmath1030 , and @xmath1031 , @xmath1032 , and @xmath1033 sr are the solid angles of beams in q , v , and w bands , respectively @xcite . using the radio sources observed in clusters of galaxies by @xcite
, @xcite estimated that the mean flux of sources in q band is 10.4 mjy , and that at 90 ghz ( which is close to 94 ghz of w band ) is @xmath1034 to 6 mjy . using these estimates
, we expect the source contamination at the level of @xmath582 to 2 @xmath1035k in v and w bands , which is @xmath1036 to 10% of the measured sz temperature . therefore , the best - fitting amplitudes reported in table [ tab : szamplitude ] could be underestimated by @xmath1036 to 10% .
the gas pressure profile is not the only factor that determines the sz power spectrum .
the other important factor is the mass function , @xmath1037 : @xmath1038 is the 2d fourier transform of @xmath1039 .
therefore , a lower - than - expected @xmath1040 may imply either a lower - than - expected amplitude of matter density fluctuations , i.e. , @xmath39 , or a lower - than - expected gas pressure , or both .
as the predictions for the sz power spectrum available today ( see , e.g. , * ? ? ?
* ; * ? ? ?
* and references therein ) are similar to the prediction of @xcite ( for example , @xcite found @xmath1041 for the prediction of @xcite , which is based on the gas model of @xcite ) , a plausible explanation for a lower - than - expected @xmath1040 is a lower - than - expected gas pressure .
@xcite find that the x - ray observed integrated pressure enclosed within @xmath774 , @xmath1042 , for a given @xmath844 is about a factor of 0.75 times the prediction from the cooling+sf simulation of @xcite .
this is in good agreement with our corresponding result for the `` high @xmath954 '' samples , @xmath996 ( 68% cl ; statistical error only ) .
while the ks profile is generally in good agreement with arnaud et al.s profile , the former is more extended than the latter ( see figure [ fig : profile ] ) , which makes the ks prediction for the projected sz profiles bigger .
note , however , that the outer slope of the fitting formula given by @xcite ( equation ( [ eq : gnfw ] ) ) has been forced to match that from hydrodynamical simulations of @xcite in @xmath1043 .
see the bottom panels of figure [ fig : profile ] .
the steepness of the profile at @xmath1044 from the simulation may be attributed to a significant non - thermal pressure support from @xmath1045 , which makes it possible to balance gravity by less thermal pressure at larger radii . in other words ,
the _ total _ pressure ( i.e. , thermal plus @xmath1046 ) profile would probably be closer to the ks prediction , but the thermal pressure would decline more rapidly than the total pressure would . if the sz effect seen in the _ wmap _ data is less than theoretically expected , what would be the implications ?
one possibility is that protons and electrons do not share the same temperature .
the electron - proton equilibration time is longer than the hubble time at the virial radius , so that the electron temperature may be lower than the proton temperature in the outer regions of clusters which contribute a significant fraction of the predicted sz flux @xcite .
the other sources of non - thermal pressure support in outskirts of the cluster ( turbulence , magnetic field , and cosmic rays ) would reduce the thermal sz effect relative to the expectation , if these effects are not taken into account in modeling the intracluster medium .
heat conduction may also play some role in suppressing the gas pressure @xcite . in order to explore the impact of gas pressure at @xmath1047
, we cut the pressure profile at @xmath982 ( instead of @xmath1048 ) and repeat the analysis .
we find @xmath1049 and @xmath985 for high and low @xmath954 clusters , respectively .
( we found @xmath1013 and @xmath973 for @xmath883 .
see table [ tab : szamplitude ] . )
these results are somewhat puzzling - the x - ray observations directly measure gas out to @xmath774 , and thus we would expect to find @xmath1050 at least out to @xmath774 . this result may suggest that , as we have shown in section [ sec : nearby ] , the problem is not with the outskirts of the cluster , but with the inner parts where the cooling flow has the largest effect .
the _ relative _ amplitudes between high and low @xmath954 clusters suggest that a significant amount of pressure is missing in low mass ( @xmath1051 ) clusters , even if we scale all the results such that high - mass clusters are forced to have @xmath887 .
a similar trend is also seen in figure 3 of @xcite .
this interpretation is consistent with the sz power spectrum being lower than theoretically expected .
the spt measures the sz power spectrum at @xmath1052 . at such high multipoles ,
the contributions to the sz power spectrum are dominated by relatively low - mass clusters , @xmath1051 ( see figure 6 of * ? ? ?
therefore , a plausible explanation for the lower - than - expected sz power spectrum is a missing pressure ( relative to theory ) in lower mass clusters . scaling relations , gas pressure , and entropy of low - mass clusters and groups
have been studied in the literature .
@xcite obtained a relation between @xmath954 of 206 x - ray - selected galaxy groups and the mass ( @xmath1053 ) derived from the stacking analysis of weak lensing measurements . converting their best - fitting relation to @xmath1003@xmath954 relation , we find @xmath1054^{0.22}$ ] .
( note that the pivot luminosity of the original scaling relation is @xmath1055 . ) as @xmath1056 , their relation is @xmath1057 higher than the fiducial scaling relation that we adopted ( equation ( [ eq : r500 ] ) ) .
had we used their scaling relation , we would find even lower normalizations .
the next generation of simulations or analytical calculations of the sz effect should be focused more on understanding the gas pressure profiles , both the amplitude and the shape , especially in low - mass clusters .
new measurements of the sz effect toward many individual clusters with unprecedented sensitivity are now becoming available @xcite . these new measurements would be important for understanding the gas pressure in low - mass clusters .
, and the tensor - to - scalar ratio , @xmath395 , derived from the data combination of _ wmap_+bao+@xmath0 .
the symbols show the predictions from `` chaotic '' inflation models whose potential is given by @xmath1058 @xcite , with @xmath1059 ( solid ) and @xmath1060 ( dashed ) for single - field models , and @xmath1060 for multi - axion field models with @xmath1061 ( dotted ; * ? ? ?
, width=302 ] with the _ wmap _
7-year temperature and polarization data , new measurements of @xmath0 @xcite , and improved large - scale structure data @xcite , we have been able to rigorously test the standard cosmological model .
the model continues to be an exquisite fit to the existing data .
depending on the parameters , we also use the other data sets such as the small - scale cmb temperature power spectra ( * ? ? ?
* ; * ? ? ?
* for the primordial helium abundance ) , the power spectrum of lrgs derived from _ sdss _ ( * ? ? ?
* for neutrino properties ) , the type ia supernova data ( * ? ? ?
* for dark energy ) , and the time - delay distance to the lens system b1608 + 656 ( * ? ? ?
* for dark energy and spatial curvature ) .
the combined data sets enable improved constraints over the _
wmap_-only constraints on the cosmological parameters presented in @xcite on physically - motivated extensions of the standard model .
we summarize the most significant findings from our analysis ( also see table [ tab : deviation ] , [ tab : yhe ] , and [ tab : darkenergy ] ) : * * gravitational waves and primordial power spectrum*. our best estimate of the spectral index of a power - law primordial power spectrum of curvature perturbations is @xmath1062 ( 68% cl ) .
we find no evidence for tensor modes : the 95% cl limit is @xmath62 .
the 5-year _
wmap_+bao+sn limit was @xmath1063 ( 95% cl ) . for comparison , the 7-year _
wmap_+bao+sn limit is @xmath61 ( 95% cl ) .
these limits do not include systematic errors in the supernova data .
] there is no evidence for the running spectral index , @xmath1064 ( 68% cl ) .
given that the improvements on @xmath29 , @xmath395 , and @xmath388 from the 5-year results are modest , their implications for models of inflation are similar to those discussed in section 3.3 of @xcite .
also see @xcite , @xcite and @xcite for more recent surveys of implications for inflation . in figure
[ fig : nsr ] , we compare the 7-year _ wmap_+bao+@xmath0 limits on @xmath29 and @xmath395 to the predictions from inflation models with monomial potential , @xmath1058 . * * neutrino properties*. better determinations of the amplitude of the third acoustic peak of the temperature power spectrum and @xmath0 have led to improved limits on the total mass of neutrinos , @xmath3 , and the effective number of neutrino species , @xmath4 ( 68% cl ) , both of which are derived from
_ wmap_+bao+@xmath0 without any information on the growth of structure .
when bao is replaced by the lrg power spectrum , we find @xmath525 , and the effective number of neutrino species , @xmath558 ( 68% cl ) .
* * primordial helium abundance*. by combining the _ wmap _ data with the small - scale cmb data , we have detected , by more than @xmath588 , a change in the silk damping on small angular scales ( @xmath1065 ) due to the effect of primordial helium on the temperature power spectrum .
we find @xmath6 ( 68% cl ) .
the astrophysical measurements of helium abundance in stars or hii regions provide tight upper limits on @xmath17 , whereas the cmb data can be used to provide a lower limit . with a conservative hard prior on @xmath593
, we find @xmath594 ( 68% cl ) .
our detection of helium at @xmath1066 contradicts versions of the `` cold big bang model , '' where most of the cosmological helium is produced by the first generation of stars @xcite .
* * parity violation*. the 7-year polarization data have significantly improved over the 5-year data .
this has led to a significantly improved limit on the rotation angle of the polarization plane due to potential parity - violating effects .
our best limit is @xmath1067 ( 68% cl ) . * * axion dark matter*. the 7-year
_ wmap_+bao+@xmath0 limit on the non - adiabatic perturbations that are uncorrelated with curvature perturbations , @xmath1068 , constrains the parameter space of axion dark matter in the context of the misalignment scenario .
it continues to suggest that a future detection of tensor - to - scalar ratio , @xmath395 , at the level of @xmath480 would require a fine - tuning of parameters such as the misalignment angle , @xmath1069 , a significant amount of entropy production between the qcd phase transition and the big bang nucleosynthesis , @xmath1070 , a super - planckian axion decay constant , @xmath1071 gev , an axion contribution to the matter density of the universe being totally sub - dominant , or a combination of all of the above with less tuning in each ( also see section 3.6.3 of * ? ? ?
the 7-year _
wmap_+bao+@xmath0 limit on correlated isocurvature perturbations , which is relevant to the curvaton dark matter , is @xmath1072 . *
* dark energy*. with _
but without high - redshift type ia supernovae , we find @xmath5 ( 68% cl ) for a flat universe .
adding the supernova data reduces the error bar by about a half . for a curved universe ,
addition of supernova data reduces the error in @xmath89 dramatically ( by a factor of more than four ) , while the error in curvature is well constrained by _ wmap_+bao+@xmath0 . in figure
[ fig : w0wa ] , we show the 7-year limits on a time - dependent equation of state in the form of @xmath1073 .
we find @xmath1074 and @xmath1075 ( 68% cl ) from _ wmap_+bao+@xmath0+sn .
the data are consistent with a flat universe dominated by a cosmological constant . *
* primordial non - gaussianity*. the 95% cl limits on physically - motivated primordial non - gaussianity parameters are @xmath84 , @xmath85 , and @xmath86 .
when combined with the limit on @xmath684 from _ sdss _ , @xmath739 @xcite , we find @xmath740 .
the data are consistent with gaussian primordial curvature perturbations . *
* sunyaev zeldovich effect*. using the optimal estimator , we have measured the sz effect towards clusters of galaxies .
we have detected the sz effect toward the coma cluster at @xmath741 , and made the statistical detection of the sz effect by optimally stacking the _ wmap _ data at the locations of known clusters of galaxies . by stacking 11 nearby massive clusters
, we detect the sz effect at @xmath743 , and find that the measured sz signal and the predictions from the x - ray data agree well . on the other hand , we find that the sz signal from the stacking analysis is about @xmath1076 times the predictions from the current generation of analytical calculations , hydrodynamical simulations , and the `` universal pressure profile '' of @xcite .
we detect the expected sz signal in relaxed clusters that have cool cores .
we find that the sz signal from non - relaxed clusters have sz signals that are 50% of the signal predicted by arnaud et al.s profile .
the discrepancy with theoretical predictions presents a puzzle .
this lower - than - theoretically - expected sz signal is consistent with the lower - than - theoretically - expected sz power spectrum recently measured by the spt collaboration @xcite . while we find a better agreement between the _ wmap _ data and the expectations for massive clusters with @xmath1077 , a significant amount of pressure ( relative to theory ) is missing in lower mass clusters .
our results imply that we may not fully understand the gas pressure in low - mass clusters .
this issue would become particularly important when the sz effect is used as a cosmological probe .
we also reported a novel analysis of the _ wmap _ temperature and polarization data that enable us to directly `` see '' the imprint of adiabatic scalar fluctuations in the maps of polarization directions around temperature hot and cold spots .
these give a striking confirmation of our understanding of the physics at the decoupling epoch in the form of radial and tangential polarization patterns at two characteristic angular scales that are important for the physics of acoustic oscillation : the compression phase at @xmath254 and the reversal phase at @xmath255 .
the cmb data have provided us with many stringent constraints on various properties of our universe .
one of many lessons that we have learned from the cmb data is that , given the data that we have , inventions of new , physically - motivated , observables beyond the spherically - averaged power spectrum often lead to new insights into the physics of the universe .
well - studied examples include primordial non - gaussianity parameters ( @xmath19 from the bispectrum ) , parity - violation angle ( @xmath284 from the tb and eb correlations ) , modulated primordial power spectrum ( @xmath1078 from direction - dependent power spectra ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* see bennett et al .
2010 for the 7-year limits ) .
the data continue to improve , including more integration of the _ wmap_observations . at the same rate ,
it is important to find more ways to subject the data to various properties of the universe that have not been explored yet .
the _ wmap _ mission is made possible by the support of the science mission directorate office at nasa headquarters .
this research was additionally supported by nasa grants nng05ge76 g , nnx07al75 g s01 , ltsa03 - 000 - 0090 , atpnng04gk55 g , adp03 - 0000 - 092 , and nnx08al43 g , and nsf grants ast-0807649 and phy-0758153 .
ek acknowledges support from an alfred p. sloan research fellowship .
jd is partly supported by an research councils uk ( rcuk ) fellowship .
we thank mike greason for his help on the analysis of the _ wmap _ data , and t. b. griswold for the artwork .
we thank d. jeong for his help on calculating the peak bias presented in section [ sec : pol ] , b. a. reid for discussion on the treatment of massive neutrinos in the expansion rate which has led to our exact treatment in section [ sec : exactnu ] , a. g. riess for discussion on the type ia supernova data set and the @xmath0 measurement , m. sullivan for discussion on the type ia supernova data set , s. h. suyu and p. j. marshall for providing us with the likelihood function for the time - delay distance and discussion on strong lensing measurements , a. vikhlinin for the x - ray data on his nearby cluster samples ( used in section [ sec : nearby ] ) , n. afshordi , p. bode , r. lieu , y .- t .
lin , d. nagai , n. sehgal , l. shaw and h. trac for discussion and feedback on section [ sec : sz ] ( sz effect ) , and f. takahashi for his help on refining the results on axion dark matter presented in section [ sec : ad ] . computations for the analysis of non - gaussianity in section [ sec : ng ] were carried out by the terascale infrastructure for groundbreaking research in engineering and science ( tigress ) at the princeton institute for computational science and engineering ( picscie ) .
this research has made use of nasa s astrophysics data system bibliographic services .
we acknowledge use of the healpix @xcite , camb @xcite , and cmbfast @xcite packages .
the constraints on the cosmological parameters reported in the original version of this paper were based on a version of camb which used a recombination history calculated by the recfast version 1.4.2 @xcite . shortly after the submission of the original version ,
a new version camb was released with the recfast version 1.5 .
this revision incorporates the improved treatment of the hydrogen and helium recombination , following numerous work done over the last several years ( see * ? ? ?
* and references therein ) .
specifically , the code multiplies the ionization fraction , @xmath1079 , by a cosmology - independent `` fudge function , '' @xmath1080 , found by @xcite . a change in the recombination history
mostly affects the time and duration of the photon decoupling which , in turn , affects the amount of silk damping .
therefore , it is expected to affect the cosmological parameters such as @xmath29 and @xmath586 @xcite . in order to see the effects of the improved recombination code on the cosmological parameters , we have re - run the @xmath1cdm chain with the latest camb code that includes recfast version 1.5 .
we find that the effects are small , and in most cases negligible compared to the error bars ; however , we find that the significance at which @xmath1081 is excluded is no longer more than @xmath588 : with the improved recombination code , we find @xmath2 ( 68% cl ) , and @xmath1081 is excluded at 99.5% cl .
finally , the _ wmap _ likelihood code has also changed from the initial version ( 4.0 ) , which used a temperature power spectrum with a slightly incorrect estimate for the residual point - source amplitude , and a te power spectrum with a slightly incorrect @xmath1082 factor . the new version ( 4.1 ) corrects both errors ; however , the change in the parameters is largely driven by the above modification of the recombination history . throughout the main body of this paper
, we have adopted the new parameters for the simplest 6-parameter @xmath1cdm model , but we have kept the previous parameters for all the other models because the changes are too small to report .
we compare the @xmath1cdm parameters derived from _
wmap_+bao+@xmath0 in table [ tab : comparison ] .
see @xcite for the comparison of _ wmap_-only parameters .
lccccc primary & @xmath20 & 2.253 & 2.246 & @xmath22 & @xmath1083 & @xmath23 & 0.1122 & 0.1120 & @xmath25 & @xmath1084 & @xmath26 & 0.728 & 0.728 & @xmath28 & @xmath1085 & @xmath29 & 0.967 & 0.961 & @xmath31 & @xmath1086 & @xmath32 & 0.085 & 0.087 & @xmath34 & @xmath1087 & @xmath1088 & @xmath36 & @xmath1089 & @xmath38 & @xmath1090 derived & @xmath39 & 0.810 & 0.807 & @xmath41 & @xmath1091 & @xmath0 & @xmath43 & @xmath1092 & @xmath45 km / s / mpc & @xmath1093 & @xmath46 & 0.0455 & 0.0455 & @xmath48 & @xmath1094 & @xmath49 & 0.226 & 0.227 & @xmath51 & @xmath1095 & @xmath52 & 0.1347 & 0.1344 & @xmath54 & @xmath1096 & @xmath55 & 10.3 & 10.5 & @xmath56 & @xmath1097 & @xmath57 & 13.76 gyr & 13.78 gyr & @xmath59 gyr & @xmath1098 [ tab : comparison ]
in order to calculate the stacked profiles of temperature and polarization of the cmb at the locations of temperature peaks , we need to relate the peak number density contrast , @xmath1099 , to the underlying temperature fluctuation , @xmath1100 .
one often encounters a similar problem in the large - scale structure of the universe : how can we relate the number density contrast of galaxies to the underlying matter density fluctuation ?
it is often assumed that the number density contrast of peaks with a given peak height @xmath746 is simply proportional to the underlying density field .
if one adopted such a linear and _ scale - independent _ bias prescription , one would find@xmath1101 . ]
@xmath1102 however , our numerical simulations show that the linear bias does not give an accurate description of @xmath144 or @xmath1103 .
in fact , breakdown of the linear bias is precisely what is expected from the statistics of peaks . from detailed investigations of the statistics of peaks , @xcite found the following _ scale - dependent _ bias : @xmath1104 \delta t(\hat{\mathbf n } ) .
\label{eq : bias}\ ] ] while the first , constant term @xmath1105 has been known for a long time @xcite , the second term @xmath1106 has been recognized only recently .
the presence of @xmath1106 is expected because , to define peaks , one needs to use the information on the first and second derivatives of @xmath1107 . as the first derivative must vanish at the locations of peaks ,
the above equation does not contain the first derivative .
@xcite has derived the explicit forms of @xmath1105 and @xmath1106 : @xmath1108 where @xmath1109 , @xmath1110 , @xmath1111 is the r.m.s . of @xmath1112th derivatives of the temperature fluctuation : @xmath1113^jc_l^{\rm tt}(w_l^t)^2,\ ] ] and @xmath151 is the harmonic transform of a window function ( which is a combination of the experimental beam , pixel window , and any other additional smoothing applied to the temperature data ) .
the quantity @xmath1114 is called the `` mean curvature , '' and is given by @xmath1115 , where @xmath1116}{\sqrt{2\pi(1-\gamma^2)}}.\ ] ] while @xcite applied this formalism to a 3-dimensional gaussian random field , it is straightforward to generalize his results to a 2-dimensional case , for which @xmath1117 is given by @xcite , @xmath1118 with the bias given by equation ( [ eq : bias ] ) , we find @xmath1119 + c_l^{\rm tb}\sin[2(\phi-\varphi)]\right\}e^{i{\mathbf l}\cdot{\bm{\theta}}},\\ \nonumber & & \langle \delta_{\rm pk}(\hat{\mathbf{n}})u_r(\hat{\mathbf{n}}+\hat{\bm{\theta}})\rangle = -\int\frac{d^2{\mathbf l}}{(2\pi)^2 } w_l^tw_l^p(b_\nu+b_\zeta l^2)\\ & & \times \left\{c_l^{\rm te}\sin[2(\phi-\varphi ) ] -c_l^{\rm tb } \cos[2(\phi-\varphi)]\right\}e^{i{\mathbf l}\cdot{\bm{\theta}}},\end{aligned}\ ] ] where @xmath151 and @xmath152 are spherical harmonic transforms of the smoothing functions applied to the temperature and polarization data , respectively . recalling @xmath1120 ,
@xmath1121e^{ix\cos(\phi-\varphi)}=0 $ ] , and @xmath1122 with @xmath1123 , @xmath1124 and @xmath1125 , we find @xmath1126 using these results in equations ( [ eq : qr1 ] ) and ( [ eq : ur1 ] ) , we finally obtain the desired formulae for the stacked polarization profiles : @xmath1127 incidentally , the stacked profile of the temperature fluctuation can also be calculated in a similar way : @xmath1128 how can we evaluate equations ( [ eq : qrapp])([eq : tapp ] ) ?
one may follow the following steps : * compute @xmath1129 , @xmath1130 , and @xmath1131 from equation ( [ eq : sigma ] ) .
for example , the _ wmap _ 5-year best - fitting temperature power spectrum for a power - law @xmath1cdm model @xcite , multiplied by a gaussian smoothing of @xmath192 full - width - at - half - maximum ( fwhm ) and the pixel window function for the healpix resolution of @xmath1132 , gives @xmath1133 , @xmath1134 , and @xmath1135 . * compute @xmath1136 . for the above example , we find @xmath1137 . * as we need to integrate over peak heights @xmath746 , we need to compute the functions @xmath1138 and @xmath1139 for various values of @xmath746 .
the former function , @xmath1138 , can be found analytically ( see equation ( a1.9 ) of * ? ? ?
* ) . for @xmath1140
, we need to integrate equation ( [ eq : gint ] ) numerically . *
compute @xmath1141 . for the above example
, we find @xmath1142 , 1.831 , 3.206 , and 5.579 for @xmath834 , 1 , 5 , and 10 , respectively .
* choose a threshold peak height @xmath1143 , and compute the mean surface number density of peaks , @xmath1144 , from equation ( a1.9 ) of @xcite : @xmath1145 the integration boundary is taken from @xmath1146 to @xmath1147 for temperature hot spots , and from @xmath1148 to @xmath1149 for temperature cold spots . for the above example
, we find @xmath1150 , @xmath1151 , @xmath1152 , and @xmath1153 for @xmath200 , 1 , 2 , and 3 , respectively . *
compute @xmath1105 and @xmath1106 from equation ( [ eq : b1b2 ] ) for various values of @xmath746 . *
average @xmath1105 and @xmath1106 over @xmath746 .
we calculate the averaged bias parameters , @xmath1154 and @xmath1155 , by integrating @xmath1105 and @xmath1106 multiplied by the number density of peaks for a given @xmath746 .
we then divide the integral by the mean number density of peaks , @xmath1144 , to find @xmath1156 the integration boundary is taken from @xmath1146 to @xmath1147 for temperature hot spots , and from @xmath1148 to @xmath1149 for temperature cold spots . for the above example , we find @xmath1157 , @xmath1158 , ( @xmath1159 , @xmath1160 ) , ( @xmath1161 , @xmath1162 ) , and ( @xmath1163 , @xmath1164 ) for @xmath1165 , 1 , 2 , and 3 , respectively ( all in units of @xmath1166 ) .
the larger the peak height is , the larger the linear bias and the smaller the scale - dependent bias becomes .
* use @xmath1154 and @xmath1155 in equations ( [ eq : qrapp ] ) and ( [ eq : urapp ] ) to compute @xmath154 and @xmath155 for a given set of @xmath145 and @xmath147 , respectively .
very roughly speaking , the bias takes on the following values : @xmath1167 & ( \mbox{for}~\nu_t=0)\\ \frac{3}{100~\mu{\rm k } } \left[1+\left(\frac{l}{290}\right)^2\right ] & ( \mbox{for}~\nu_t=3 ) \end{cases}.\end{aligned}\ ] ] the scale dependence of bias becomes important at @xmath1168 for @xmath200 , but the higher peaks are closer to having a linear bias on large scales .
one may also rewrite this using the stacked temperature values at the center , @xmath1169 , @xmath1170 , @xmath1171 , @xmath1172 for @xmath200 , 1 , 2 , and 3 : @xmath1173.\ ] ]
in this appendix , we describe an optimal likelihood - based method for estimating the stacked sz profile around clusters whose locations are taken from external catalogs .
formally , we can set up the problem as follows .
we represent the _ wmap _ data as a vector of length @xmath1174 and denote it by @xmath785 , where the index @xmath1175 ranges over channels , and the index @xmath1176 ranges over sky pixels .
( we typically take @xmath1177 corresponding to v1 , v2 , w1 , w2 , w3 , and w4 ; and @xmath1178 corresponding to a healpix resolution of @xmath796 . )
we model the _ wmap _ data as a sum of cmb , noise , and sz contributions as follows : @xmath1179 in this equation , we have written the sz contribution as a sum of @xmath1180 template maps , @xmath1181 , whose coefficients @xmath1182 are free parameters to be determined . the operator , @xmath1183 , in eq .
( [ eq : sz_dsplit ] ) converts a harmonic - space cmb realization , @xmath1184 , into a set of maps with black - body frequency dependence and channel - dependent beam convolution .
more formally , the matrix element @xmath1183 is defined by @xmath1185 where @xmath1186 is the beam transfer function ( including healpix window function ) for channel @xmath746 .
the specific form of the template maps , @xmath1187 , will depend on the type of profile reconstruction which is desired .
for example , if we want to estimate a stacked amplitude for the sz signal in @xmath1188 angular bins , we define one template for each bin .
if the bin corresponds to angular range @xmath1189 , we define a map @xmath1190 which is = 1 if the angular distance @xmath127 between pixel @xmath786 and _ some _ galaxy cluster in the catalog is in the range @xmath1189 , and zero otherwise .
we convolve this map with the beam in each channel @xmath746 and multiply by the sz frequency dependence to obtain the template @xmath814 . as another example , if we want to fit for an overall multiple of a fiducial model @xmath1190 for the total sz signal ( summed over all clusters ) then we define a single ( i.e. @xmath1191 ) template @xmath814 by applying beam convolution and the sz frequency dependence for each channel @xmath746 .
given this setup , we would like to compute the likelihood function @xmath1192 $ ] for the profile @xmath1182 , given the noisy data @xmath785 , marginalizing over the cmb realization .
we assume a fixed fiducial model @xmath1193 and represent the cmb signal covariance by a ( diagonal ) matrix @xmath1194 in harmonic space : @xmath1195 we represent the noise covariance by an ( also diagonal ) pixel - matrix @xmath1196 : @xmath1197 the joint ( cmb , sz ) likelihood function can now be written ( up to an overall normalizing constant ) : @xmath1198 \propto \exp\left [ -\frac{1}{2 } a^t s_{\rm harm}^{-1 } a - \frac{1}{2 } ( d - aa - p_\alpha t_\alpha)^t n^{-1 } ( d - aa - p_\alpha t_\alpha ) \right ] .
\label{eq : likelihood_sz}\ ] ] ( in this equation we have omitted some indexes for notational compactness , e.g. @xmath1199 and @xmath1200 .
the summation over @xmath782 is assumed . )
we can now integrate out the cmb realization @xmath884 to obtain the marginalized likelihood for the profile : @xmath1201 & = & \int da\ , \l[a , p|d ] \\ & \propto & \exp\left [ -\frac{1}{2 } ( p_\alpha - \hp_\alpha)^t f_{\alpha\beta } ( p_\beta - \hp_\beta ) \right ] , \label{eq : sz_p_d}\end{aligned}\ ] ] where we have defined the @xmath1202-by-@xmath1202 matrix @xmath1203^{-1}_{\nu'p',\nu p } ( t_\beta)_{\nu p } , \label{eq : f_def}\ ] ] and the length-@xmath1202 vector @xmath1204^{-1}_{\nu'p',\nu p } d_{\nu p}. \label{eq : hp_def}\ ] ] the likelihood function @xmath1205 $ ] in eq .
( [ eq : sz_p_d ] ) has a simple interpretation .
the likelihood for @xmath1182 is a gaussian with mean @xmath1206 and covariance matrix @xmath1207 .
this is the main result of this section , and is the basis for all our sz results in the body of the paper .
for example , when we reconstruct the stacked sz profile in angular bins , the estimated profile in each bin @xmath782 is given by @xmath1206 and the @xmath793 error is given by @xmath794 .
it is worth mentioning that the statistic @xmath1206 also appears naturally if we use an estimator framework rather than a likelihood formalism . if we think of @xmath1206 , defined by eq .
( [ eq : hp_def ] ) , as an estimator for the profile given the data @xmath1208 , then one can verify that the estimator is unbiased ( i.e. @xmath1209 , where the expectation value is taken over random cmb + noise realizations with a fixed sz contribution ) and its covariance is @xmath1207 .
conversely , it is not hard to show that @xmath1206 is the unbiased estimator with minimum variance , thus obtaining @xmath1206 in a different way .
this alternate derivation also shows that the error bars on the profile obtained in our likelihood formalism are the same as would be obtained in a direct monte carlo treatment . either from the likelihood or estimator formalism
, one sees that the statistic @xmath1206 is optimal . in the limit where all frequency channels are in the rayleigh - jeans regime , the statistic @xmath1206 corresponds to @xmath720-filtering the data and multiplying by each template map . in this case , the @xmath720 filter acts as a high - pass filter which optimally suppresses cmb power on scales larger than the clusters , and also optimally weights the channels ( in a way which is @xmath1210-dependent if the beams differ ) .
when channels with higher frequency are included , the statistic @xmath1206 would get most of its information from linear combinations of channels which contain zero cmb signal , but nonzero response to an sz signal .
( such a combination of channels does not need to be high - pass filtered , increasing its statistical weight . ) for the v+w combination in _ wmap _ , _ the @xmath801-filtered ( v@xmath231w ) null map is used to separate the sz effect and cmb _ , as cmb is canceled in this map while the sz is effect not .
we conclude with a few comments on implementation .
inspection of eqs .
( [ eq : f_def ] ) and ( [ eq : hp_def ] ) shows that it would be straightforward to compute @xmath792 and @xmath1206 , given an algorithm for multiplying a set of @xmath1211 pixel space maps @xmath785 by the operator @xmath1212^{-1}$ ] . a fast multigrid - based algorithm for this inverse problem
was found in @xcite but there is one small wrinkle in the implementation : in @xcite the problem was formulated in harmonic space and an algorithm was given for multiplying by the operator @xmath1213^{-1}$ ] .
however , the matrix identity @xmath1214^{-1 } = n_{\rm pix}^{-1 } - a [ s_{\rm harm}^{-1 } + a^t n_{\rm pix}^{-1 } a ] ^{-1 } a^t , \label{eq : sz_resum}\ ] ] allows us to relate the two inverse problems .
in fact there is another advantage to using the expression on the rhs of eq .
( [ eq : sz_resum ] ) : because the inverse noise , @xmath1215 , appears instead of the noise covariance @xmath1216 , a galactic mask can be straightforwardly included in the analysis by zeroing the matrix entries of @xmath1215 which correspond to masked pixels , so that the pixels are treated as infinitely noisy .
recently , @xcite found that the following parametrized phenomenological electron pressure profile , which is based on a `` generalized navarro - frenk - white profile '' proposed by @xcite , fits the electron pressure profiles directly derived from x - ray data of clusters well ( see equation ( 13 ) of * ? ? ? * ) : @xmath1217^{2/3+\alpha_p } p(r / r_{500}),\end{aligned}\ ] ] where @xmath1218 , @xmath957^{1/2}$ ] for a @xmath1cdm model , @xmath774 is the radius within which the mean overdensity is 500 times the critical density of the universe , @xmath1219 , and @xmath844 is the mass enclosed within @xmath774 : @xmath1220r_{500}^3 .
\label{eq : m500}\ ] ] the function @xmath1221 is defined by @xmath1222^{(\beta-\gamma)/\alpha } } , \label{eq : gnfw}\ ] ] where @xmath1223 , @xmath1224 , @xmath1225 , and @xmath1226 .
the spt collaboration stacked the sz maps of 11 known clusters and fitted the stacked sz radial profile to the above form , finding @xmath1227 , @xmath1228 , @xmath1229 , and @xmath1230 @xcite . while they did not compare the overall amplitude ( which is the focus of our analysis ) , the _ shape _ of the pressure profile found by the spt collaboration ( using the sz data ) is in an excellent agreement with that found by @xcite ( using the x - ray data ) .
the ks profile builds on and extends the idea originally put forward by @xcite and @xcite : ( i ) gas is in hydrostatic equilibrium with gravitational potential given by a navarro - frenk - white ( nfw ) dark matter density profile @xcite and ( ii ) the equation of state of gas is given by a polytropic form , @xmath1231 .
however , this model still contains two free parameters : a polytropic index @xmath481 and the normalization of @xmath1232 .
@xcite found that an additional , physically reasonable assumption that ( iii ) the slope of the gas density profile and that of the dark matter density profile agree at around the virial radius , uniquely fixes @xmath481 , leaving only one free parameter : the normalization of @xmath1232 .
these assumptions are supported by hydrodynamical simulations of clusters of galaxies , and the resulting shape of the ks profile indeed agrees with simulations reasonably well ( see , however , section [ sec : sz_discussion ] for a discussion on the shape of the profile in the outer parts of clusters ) . determining the normalization of the ks profile requires an additional assumption , described below .
also note that this model does not take into account any non - thermal pressure ( such as @xmath1046 where @xmath1233 is the bulk or turbulent velocity ) , gas cooling , or star formation ( see , e.g. , * ? ? ?
* ; * ? ? ?
* and references therein for various attempts to incorporate more gas physics ) .
the ks _ gas _ pressure profile is given by ( see section 3.3 of * ? ? ?
* for more detailed descriptions ) @xmath1234^\gamma.\ ] ] the _ electron _ pressure profile , @xmath1235 , is then given by @xmath1236p_{\rm gas}=0.518p_{\rm gas}$ ] for @xmath931 . here , @xmath1237 is the so - called `` scale radius '' of the nfw profile , and a function @xmath1238 is defined by @xmath1239 \right\}^{1/(\gamma-1)},\ ] ] with @xmath1240^{-1},\ ] ] @xmath1241 and @xmath1242 here , @xmath752 is the so - called `` concentration parameter '' of the nfw profile , which is related to the scale radius , @xmath1237 , via @xmath1243 , and @xmath1244 is the virial radius .
the virial radius gives the virial mass , @xmath938 , as @xmath1245r_{\rm vir}^3 .
\label{eq : mvir}\ ] ] here , @xmath1246 depends on @xmath1247 and @xmath26 as @xcite @xmath1248 - 39[\omega(z)-1]^2,\ ] ] where @xmath1249 ( also see * ? ? ? * ; * ? ? ?
* for other fitting formulae ) . for @xmath1250
, one finds @xmath1251 . the central gas pressure , @xmath1252 , is given by @xmath1253 \left[\frac{k_bt_{\rm gas}(0)}{8~{\rm kev}}\right],\ ] ] where @xmath1254 is the the boltzmann constant .
the central gas temperature , @xmath1255 , is given by @xmath1256.\ ] ] the central gas density , @xmath1257 , will be determined such that the gas density at the virial radius is the cosmic mean baryon fraction , @xmath936 , times the dark matter density at the same radius .
_ this is an assumption .
_ in fact , the cosmic mean merely provides an upper limit on the baryon fraction of clusters , and thus we expect the gas pressure to be less than what is given here .
how much less needs to be determined from observations ( or possibly from more detailed modeling of the intracluster medium ) . in any case , with this assumption , we find @xmath1258 ^ 3}\\ & & \times \frac{c^2}{(1+c)^2}\frac1{y_{\rm gas}(c ) } \left[\ln(1+c)-\frac1{1+c}\right]^{-1}.\end{aligned}\ ] ] this equation fixes a typo in equation ( 21 ) of @xcite . the virial radius , @xmath1244 , is approximately given by @xmath1259 ; thus , @xmath938 is approximately given by @xmath1260
. however , the exact relation depends on the mass ( see , e.g. , figure 1 of * ? ? ? * ) .
we calculate the mass within a given radius , @xmath395 , by integrating the nfw density profile @xcite : @xmath1261 specifically , for a given @xmath844 and @xmath774 , we solve the following non - linear equation for @xmath938 : @xmath1262 where @xmath1263 . here , @xmath1244 is related to @xmath938 via equation ( [ eq : mvir ] ) .
we also need a relation between the concentration parameter , @xmath752 , and @xmath1264 .
@xcite used @xmath1265 which was adopted from @xcite .
recently , @xcite ran larg e @xmath1188-body simulations with the _ wmap _
5-year cosmological parameters to find a more accurate fitting formula for the concentration parameter : @xmath1266 this formula makes clusters of galaxies ( @xmath1267 ) more concentrated than @xmath1268 would predict .
, c. l. , bay , m. , halpern , m. , hinshaw , g. , jackson , c. , jarosik , n. , kogut , a. , limon , m. , meyer , s. s. , page , l. , spergel , d. n. , tucker , g. s. , wilkinson , d. t. , wollack , e. , & wright , e. l. 2003 , , 583 , 1 , c. l. , halpern , m. , hinshaw , g. , jarosik , n. , kogut , a. , limon , m. , meyer , s. s. , page , l. , spergel , d. n. , tucker , g. s. , wollack , e. , wright , e. l. , barnes , c. , greason , m. r. , hill , r. s. , komatsu , e. , nolta , m. r. , odegard , n. , peiris , h. v. , verde , l. , & weiland , j. l. 2003 , , 148 , 1 , c. l. , hill , r. s. , hinshaw , g. , nolta , m. r. , odegard , n. , page , l. , spergel , d. n. , weiland , j. l. , wright , e. l. , halpern , m. , jarosik , n. , kogut , a. , limon , m. , meyer , s. s. , tucker , g. s. , & wollack , e. 2003 , , 148 , 97 , h. , schuecker , p. , guzzo , l. , collins , c. a. , voges , w. , cruddace , r. g. , ortiz - gil , a. , chincarini , g. , de grandi , s. , edge , a. c. , macgillivray , h. t. , neumann , d. m. , schindler , s. , & shaver , p. 2004
, , 425 , 367 , h. , schuecker , p. , pratt , g. w. , arnaud , m. , ponman , t. j. , croston , j. h. , borgani , s. , bower , r. g. , briel , u. g. , collins , c. a. , donahue , m. , forman , w. r. , finoguenov , a. , geller , m. j. , guzzo , l. , henry , j. p. , kneissl , r. , mohr , j. j. , matsushita , k. , mullis , c. r. , ohashi , t. , pedersen , k. , pierini , d. , quintana , h. , raychaudhury , s. , reiprich , t. h. , romer , a. k. , rosati , p. , sabirli , k. , temple , r. f. , viana , p. t. p. , vikhlinin , a. , voit , g. m. , & zhang , y. 2007 , , 469 , 363 , j. r. , contaldi , c. r. , pen , u .- l . , pogosyan , d. , prunet , s. , ruetalo , m. i. , wadsley , j. w. , zhang , p. , mason , b. s. , myers , s. t. , pearson , t. j. , readhead , a. c. s. , sievers , j. l. , & udomprasert , p. s. 2005 , , 626 , 12 , m. l. , ade , p. , bock , j. , bowden , m. , cahill , g. , castro , p. g. , church , s. , culverhouse , t. , friedman , r. b. , ganga , k. , gear , w. k. , gupta , s. , hinderks , j. , kovac , j. , lange , a. e. , leitch , e. , melhuish , s. j. , memari , y. , murphy , j. a. , orlando , a. , osullivan , c. , piccirillo , l. , pryke , c. , rajguru , n. , rusholme , b. , schwarz , r. , taylor , a. n. , thompson , k. l. , turner , a. h. , wu , e. y. s. , zemcov , m. , & the qua d collaboration .
2009 , , 705 , 978 , h. c. , ade , p. a. r. , barkats , d. , battle , j. o. , bierman , e. m. , bock , j. j. , dowell , c. d. , duband , l. , hivon , e. f. , holzapfel , w. l. , hristov , v. v. , jones , w. c. , keating , b. g. , kovac , j. m. , kuo , c. l. , lange , a. e. , leitch , e. m. , mason , p. v. , matsumura , t. , nguyen , h. t. , ponthieu , n. , pryke , c. , richter , s. , rocha , g. , sheehy , c. , takahashi , y. d. , tolan , j. e. , & yoon , k. w. 2010 , , 711 , 1123 , t. m. , mrtsell , e. , sollerman , j. , becker , a. c. , blondin , s. , challis , p. , clocchiatti , a. , filippenko , a. v. , foley , r. j. , garnavich , p. m. , jha , s. , krisciunas , k. , kirshner , r. p. , leibundgut , b. , li , w. , matheson , t. , miknaitis , g. , pignata , g. , rest , a. , riess , a. g. , schmidt , b. p. , smith , r. c. , spyromilio , j. , stubbs , c. w. , suntzeff , n. b. , tonry , j. l. , wood - vasey , w. m. , & zenteno , a. 2007 , , 666 , 716 , j. , komatsu , e. , nolta , m. r. , spergel , d. n. , larson , d. , hinshaw , g. , page , l. , bennett , c. l. , gold , b. , jarosik , n. , weiland , j. l. , halpern , m. , hill , r. s. , kogut , a. , limon , m. , meyer , s. s. , tucker , g. s. , wollack , e. , & wright , e. l. 2009 , , 180 , 306 , d. j. , zehavi , i. , hogg , d. w. , scoccimarro , r. , blanton , m. r. , nichol , r. c. , scranton , r. , seo , h .- j .
, tegmark , m. , zheng , z. , anderson , s. f. , annis , j. , bahcall , n. , brinkmann , j. , burles , s. , castander , f. j. , connolly , a. , csabai , i. , doi , m. , fukugita , m. , frieman , j. a. , glazebrook , k. , gunn , j. e. , hendry , j. s. , hennessy , g. , ivezi , z. , kent , s. , knapp , g. r. , lin , h. , loh , y .- s . , lupton , r. h. , margon , b. , mckay , t. a. , meiksin , a. , munn , j. a. , pope , a. , richmond , m. w. , schlegel , d. , schneider , d. p. , shimasaku , k. , stoughton , c. , strauss , m. a. , subbarao , m. , szalay , a. s. , szapudi , i. , tucker , d. l. , yanny , b. , & york , d. g. 2005 , , 633 , 560 , j. w. , acquaviva , v. , ade , p. a. r. , aguirre , p. , amiri , m. , appel , j. w. , barrientos , l. f. , battistelli , e. s. , bond , j. r. , brown , b. , burger , b. , chervenak , j. , das , s. , devlin , m. j. , dicker , s. r. , doriese , w. b. , dunkley , j. , dnner , r. , essinger - hileman , t. , fisher , r. p. , hajian , a. , halpern , m. , hasselfield , m. , hernndez - monteagudo , c. , hilton , g. c. , hilton , m. , hincks , a. d. , hlozek , r. , huffenberger , k. m. , hughes , d. h. , hughes , j. p. , infante , l. , irwin , k. d. , jimenez , r. , juin , j. b. , kaul , m. , klein , j. , kosowsky , a. , lau , j. m. , limon , m. , lin , y. , lupton , r. h. , marriage , t. a. , marsden , d. , martocci , k. , mauskopf , p. , menanteau , f. , moodley , k. , moseley , h. , netterfield , c. b. , niemack , m. d. , nolta , m. r. , page , l. a. , parker , l. , partridge , b. , quintana , h. , reid , b. , sehgal , n. , sievers , j. , spergel , d. n. , staggs , s. t. , swetz , d. s. , switzer , e. r. , thornton , r. , trac , h. , tucker , c. , verde , l. , warne , r. , wilson , g. , wollack , e. , & zhao , y. 2010 , , 722 , 1148 , w. l. , madore , b. f. , gibson , b. k. , ferrarese , l. , kelson , d. d. , sakai , s. , mould , j. r. , kennicutt , r. c. , ford , h. c. , graham , j. a. , huchra , j. p. , hughes , s. m. g. , illingworth , g. d. , macri , l. m. , & stetson , p. b. 2001 , , 553 , 47 , l. , semboloni , e. , hoekstra , h. , kilbinger , m. , van waerbeke , l. , tereno , i. , mellier , y. , heymans , c. , coupon , j. , benabed , k. , benjamin , j. , bertin , e. , dor , o. , hudson , m. j. , ilbert , o. , maoli , r. , marmo , c. , mccracken , h. j. , & mnard , b. 2008 , , 479 , 9 , b. , bennett , c. l. , hill , r. s. , hinshaw , g. , odegard , n. , spergel , d. n. , weiland , j. , dunkley , j. , halpern , m. , jarosik , n. , kogut , a. , komatsu , e. , larson , d. , meyer , s. s. , nolta , m. , wollack , e. , & wright , e. l. 2009 , , 180 , 265 , j. , astier , p. , baumont , s. , hardin , d. , pain , r. , regnault , n. , basa , s. , carlberg , r. g. , conley , a. , fabbro , s. , fouchez , d. , hook , i. m. , howell , d. a. , perrett , k. , pritchet , c. j. , rich , j. , sullivan , m. , antilogus , p. , aubourg , e. , bazin , g. , bronder , j. , filiol , m. , palanque - delabrouille , n. , ripoche , p. , & ruhlmann - kleider , v. 2007 , , 466 , 11 , m. , challis , p. , jha , s. , kirshner , r. p. , matheson , t. , modjaz , m. , rest , a. , michael wood - vasey , w. , bakos , g. , barton , e. j. , berlind , p. , bragg , a. , briceo , c. , brown , w. r. , caldwell , n. , calkins , m. , cho , r. , ciupik , l. , contreras , m. , dendy , k. , dosaj , a. , durham , n. , eriksen , k. , esquerdo , g. , everett , m. , falco , e. , fernandez , j. , gaba , a. , garnavich , p. , graves , g. , green , p. , groner , t. , hergenrother , c. , holman , m. j. , hradecky , v. , huchra , j. , hutchison , b. , jerius , d. , jordan , a. , kilgard , r. , krauss , m. , luhman , k. , macri , l. , marrone , d. , mcdowell , j. , mcintosh , d. , mcnamara , b. , megeath , t. , mochejska , b. , munoz , d. , muzerolle , j. , naranjo , o. , narayan , g. , pahre , m. , peters , w. , peterson , d. , rines , k. , ripman , b. , roussanova , a. , schild , r. , sicilia - aguilar , a. , sokoloski , j. , smalley , k. , smith , a. , spahr , t. , stanek , k. z. , barmby , p. , blondin , s. , stubbs , c. w. , szentgyorgyi , a. , torres , m. a. p. , vaz , a. , vikhlinin , a. , wang , z. , westover , m. , woods , d. , & zhao , p. 2009
, , 700 , 331 , a. d. , acquaviva , v. , ade , p. , aguirre , p. , amiri , m. , appel , j. w. , barrientos , l. f. , battistelli , e. s. , bond , j. r. , brown , b. , burger , b. , chervenak , j. , das , s. , devlin , m. j. , dicker , s. , doriese , w. b. , dunkley , j. , dnner , r. , essinger - hileman , t. , fisher , r. p. , fowler , j. w. , hajian , a. , halpern , m. , hasselfield , m. , hernndez - monteagudo , c. , hilton , g. c. , hilton , m. , hlozek , r. , huffenberger , k. , hughes , d. , hughes , j. p. , infante , l. , irwin , k. d. , jimenez , r. , juin , j. b. , kaul , m. , klein , j. , kosowsky , a. , lau , j. m. , limon , m. , lin , y. , lupton , r. h. , marriage , t. , marsden , d. , martocci , k. , mauskopf , p. , menanteau , f. , moodley , k. , moseley , h. , netterfield , c. b. , niemack , m. d. , nolta , m. r. , page , l. a. , parker , l. , partridge , b. , quintana , h. , reid , b. , sehgal , n. , sievers , j. , spergel , d. n. , staggs , s. t. , stryzak , o. , swetz , d. , switzer , e. , thornton , r. , trac , h. , tucker , c. , verde , l. , warne , r. , wilson , g. , wollack , e. , & zhao , y. 2009 , arxiv e - prints , arxiv:0907.0461 , g. , nolta , m. r. , bennett , c. l. , bean , r. , dor , o. , greason , m. r. , halpern , m. , hill , r. s. , jarosik , n. , kogut , a. , komatsu , e. , limon , m. , odegard , n. , meyer , s. s. , page , l. , peiris , h. v. , spergel , d. n. , tucker , g. s. , verde , l. , weiland , j. l. , wollack , e. , & wright , e. l. 2007 , , 170 , 288 , g. , weiland , j. l. , hill , r. s. , odegard , n. , larson , d. , bennett , c. l. , dunkley , j. , gold , b. , greason , m. r. , jarosik , n. , komatsu , e. , nolta , m. r. , page , l. , spergel , d. n. , wollack , e. , halpern , m. , kogut , a. , limon , m. , meyer , s. s. , tucker , g. s. , & wright , e. l. 2009 , , 180 , 225 , h. , fomalont , e. b. , horiuchi , s. , lovell , j. e. j. , moellenbrock , g. a. , inoue , m. , burke , b. f. , dewdney , p. e. , gurvits , l. i. , kobayashi , h. , jauncey , d. l. , murata , y. , mcculloch , p. , preston , r. a. , avruch , i. m. , edwards , p. g. , dougherty , s. m. , scott , w. k. , frey , s. , paragi , z. , kovalev , y. a. , popov , m. , romney , j. d. , schilizzi , r. t. , shen , z. , nicolson , g. , quick , j. , costa , m. , dodson , r. , reynolds , j. e. , tzioumis , a. k. , tingay , s. j. , hong , x. , liang , s. , huang , x. , wei , w. , trigilio , c. , tuccari , g. , nakajima , j. , kawai , e. , umemoto , t. , miyaji , t. , fujisawa , k. , kus , n. k. a. , ghigo , f. , salter , c. , ghosh , t. , kanevsky , b. , slysh , v. , gunn , a. , burgess , p. , carlson , b. , del rizzo , d. , taylor , r. , cannon , w. , kameno , s. , shibata , k. m. , benson , j. , flatters , c. , hale , a. , lewis , c. , langston , g. , minter , a. , miller , k. , smith , j. , wietfeldt , r. , altunin , v. , meier , d. l. , murphy , d. w. , resch , g. , lister , m. l. , piner , b. g. , jenkins , r. , border , j. , & gimeno , j. 2000 , , 52 , 997 , w. c. , ade , p. a. r. , bock , j. j. , bond , j. r. , borrill , j. , boscaleri , a. , cabella , p. , contaldi , c. r. , crill , b. p. , de bernardis , p. , de gasperis , g. , de oliveira - costa , a. , de troia , g. , di stefano , g. , hivon , e. , jaffe , a. h. , kisner , t. s. , lange , a. e. , mactavish , c. j. , masi , s. , mauskopf , p. d. , melchiorri , a. , montroy , t. e. , natoli , p. , netterfield , c. b. , pascale , e. , piacentini , f. , pogosyan , d. , polenta , g. , prunet , s. , ricciardi , s. , romeo , g. , ruhl , j. e. , santini , p. , tegmark , m. , veneziani , m. , & vittorio , n. 2006 , , 647 , 823 , r. , becker , a. c. , cinabro , d. , vanderplas , j. , frieman , j. a. , marriner , j. , davis , t. m. , dilday , b. , holtzman , j. , jha , s. w. , lampeitl , h. , sako , m. , smith , m. , zheng , c. , nichol , r. c. , bassett , b. , bender , r. , depoy , d. l. , doi , m. , elson , e. , filippenko , a. v. , foley , r. j. , garnavich , p. m. , hopp , u. , ihara , y. , ketzeback , w. , kollatschny , w. , konishi , k. , marshall , j. l. , mcmillan , r. j. , miknaitis , g. , morokuma , t. , mrtsell , e. , pan , k. , prieto , j. l. , richmond , m. w. , riess , a. g. , romani , r. , schneider , d. p. , sollerman , j. , takanashi , n. , tokita , k. , van der heyden , k. , wheeler , j. c. , yasuda , n. , & york , d. 2009 , , 185 , 32 , b. p. , mckay , t. a. , annis , j. , wechsler , r. h. , evrard , a. , bleem , l. , becker , m. , johnston , d. , sheldon , e. , nichol , r. , miller , c. , scranton , r. , bahcall , n. , barentine , j. , brewington , h. , brinkmann , j. , harvanek , m. , kleinman , s. , krzesinski , j. , long , d. , nitta , a. , schneider , d. p. , sneddin , s. , voges , w. , & york , d. 2007 , , 660 , 239 , e. , dunkley , j. , nolta , m. r. , bennett , c. l. , gold , b. , hinshaw , g. , jarosik , n. , larson , d. , limon , m. , page , l. , spergel , d. n. , halpern , m. , hill , r. s. , kogut , a. , meyer , s. s. , tucker , g. s. , weiland , j. l. , wollack , e. , & wright , e. l. 2009 , , 180 , 330 , e. , kogut , a. , nolta , m. r. , bennett , c. l. , halpern , m. , hinshaw , g. , jarosik , n. , limon , m. , meyer , s. s. , page , l. , spergel , d. n. , tucker , g. s. , verde , l. , wollack , e. , & wright , e. l. 2003 , , 148 , 119 , m. , rubin , d. , aldering , g. , agostinho , r. j. , amadon , a. , amanullah , r. , balland , c. , barbary , k. , blanc , g. , challis , p. j. , conley , a. , connolly , n. v. , covarrubias , r. , dawson , k. s. , deustua , s. e. , ellis , r. , fabbro , s. , fadeyev , v. , fan , x. , farris , b. , folatelli , g. , frye , b. l. , garavini , g. , gates , e. l. , germany , l. , goldhaber , g. , goldman , b. , goobar , a. , groom , d. e. , haissinski , j. , hardin , d. , hook , i. , kent , s. , kim , a. g. , knop , r. a. , lidman , c. , linder , e. v. , mendez , j. , meyers , j. , miller , g. j. , moniez , m. , mouro , a. m. , newberg , h. , nobili , s. , nugent , p. e. , pain , r. , perdereau , o. , perlmutter , s. , phillips , m. m. , prasad , v. , quimby , r. , regnault , n. , rich , j. , rubenstein , e. p. , ruiz - lapuente , p. , santos , f. d. , schaefer , b. e. , schommer , r. a. , smith , r. c. , soderberg , a. m. , spadafora , a. l. , strolger , l. , strovink , m. , suntzeff , n. b. , suzuki , n. , thomas , r. c. , walton , n. a. , wang , l. , wood - vasey , w. m. , & yun , j. l. 2008 , , 686 , 749 , a. , finoguenov , a. , kneib , j. , taylor , j. e. , massey , r. , rhodes , j. , ilbert , o. , bundy , k. , tinker , j. , george , m. r. , capak , p. , koekemoer , a. m. , johnston , d. e. , zhang , y. , cappelluti , n. , ellis , r. s. , elvis , m. , giodini , s. , heymans , c. , le fvre , o. , lilly , s. , mccracken , h. j. , mellier , y. , rfrgier , a. , salvato , m. , scoville , n. , smoot , g. , tanaka , m. , van waerbeke , l. , & wolk , m. 2010 , , 709 , 97 , m. , reichardt , c. l. , schaffer , k. k. , zahn , o. , ade , p. a. r. , aird , k. a. , benson , b. a. , bleem , l. e. , carlstrom , j. e. , chang , c. l. , cho , h. , crawford , t. m. , crites , a. t. , de haan , t. , dobbs , m. a. , george , e. m. , hall , n. r. , halverson , n. w. , holder , g. p. , holzapfel , w. l. , hrubes , j. d. , joy , m. , keisler , r. , knox , l. , lee , a. t. , leitch , e. m. , mcmahon , j. j. , mehl , j. , meyer , s. s. , mohr , j. j. , montroy , t. e. , padin , s. , plagge , t. , pryke , c. , ruhl , j. e. , shaw , l. , shirokoff , e. , spieler , h. g. , stalder , b. , staniszewski , z. , stark , a. a. , vanderlinde , k. , vieira , j. d. , & williamson , r. 2010 , , 719 , 1045 , s. , ade , p. a. r. , bock , j. j. , bond , j. r. , borrill , j. , boscaleri , a. , cabella , p. , contaldi , c. r. , crill , b. p. , de bernardis , p. , de gasperis , g. , de oliveira - costa , a. , de troia , g. , di stefano , g. , ehlers , p. , hivon , e. , hristov , v. , iacoangeli , a. , jaffe , a. h. , jones , w. c. , kisner , t. s. , lange , a. e. , mactavish , c. j. , marini bettolo , c. , mason , p. , mauskopf , p. d. , montroy , t. e. , nati , f. , nati , l. , natoli , p. , netterfield , c. b. , pascale , e. , piacentini , f. , pogosyan , d. , polenta , g. , prunet , s. , ricciardi , s. , romeo , g. , ruhl , j. e. , santini , p. , tegmark , m. , torbet , e. , veneziani , m. , & vittorio , n. 2006 , , 458 , 687 , r. , rhodes , j. , leauthaud , a. , capak , p. , ellis , r. , koekemoer , a. , rfrgier , a. , scoville , n. , taylor , j. e. , albert , j. , berg , j. , heymans , c. , johnston , d. , kneib , j .-
mellier , y. , mobasher , b. , semboloni , e. , shopbell , p. , tasca , l. , & van waerbeke , l. 2007 , , 172 , 239 , t. e. , ade , p. a. r. , bock , j. j. , bond , j. r. , borrill , j. , boscaleri , a. , cabella , p. , contaldi , c. r. , crill , b. p. , de bernardis , p. , de gasperis , g. , de oliveira - costa , a. , de troia , g. , di stefano , g. , hivon , e. , jaffe , a. h. , kisner , t. s. , jones , w. c. , lange , a. e. , masi , s. , mauskopf , p. d. , mactavish , c. j. , melchiorri , a. , natoli , p. , netterfield , c. b. , pascale , e. , piacentini , f. , pogosyan , d. , polenta , g. , prunet , s. , ricciardi , s. , romeo , g. , ruhl , j. e. , santini , p. , tegmark , m. , veneziani , m. , & vittorio , n. 2006 , , 647 , 813 , s. t. , fassnacht , c. d. , djorgovski , s. g. , blandford , r. d. , matthews , k. , neugebauer , g. , pearson , t. j. , readhead , a. c. s. , smith , j. d. , thompson , d. j. , womble , d. s. , browne , i. w. a. , wilkinson , p. n. , nair , s. , jackson , n. , snellen , i. a. g. , miley , g. k. , de bruyn , a. g. , & schilizzi , r. t. 1995 , , 447 , l5 + , m. r. , dunkley , j. , hill , r. s. , hinshaw , g. , komatsu , e. , larson , d. , page , l. , spergel , d. n. , bennett , c. l. , gold , b. , jarosik , n. , odegard , n. , weiland , j. l. , wollack , e. , halpern , m. , kogut , a. , limon , m. , meyer , s. s. , tucker , g. s. , & wright , e. 2009 , , 180 , 296 , l. , hinshaw , g. , komatsu , e. , nolta , m. r. , spergel , d. n. , bennett , c. l. , barnes , c. , bean , r. , dor , o. , dunkley , j. , halpern , m. , hill , r. s. , jarosik , n. , kogut , a. , limon , m. , meyer , s. s. , odegard , n. , peiris , h. v. , tucker , g. s. , verde , l. , weiland , j. l. , wollack , e. , & wright , e. l. 2007 , , 170 , 335 , l. , jackson , c. , barnes , c. , bennett , c. , halpern , m. , hinshaw , g. , jarosik , n. , kogut , a. , limon , m. , meyer , s. s. , spergel , d. n. , tucker , g. s. , wilkinson , d. t. , wollack , e. , & wright , e. l. 2003 , , 585 , 566 , h. v. , komatsu , e. , verde , l. , spergel , d. n. , bennett , c. l. , halpern , m. , hinshaw , g. , jarosik , n. , kogut , a. , limon , m. , meyer , s. s. , page , l. , tucker , g. s. , wollack , e. , & wright , e. l. 2003 , , 148 , 213 , w. j. , reid , b. a. , eisenstein , d. j. , bahcall , n. a. , budavari , t. , frieman , j. a. , fukugita , m. , gunn , j. e. , ivezi , . , knapp , g. r. , kron , r. g. , loveday , j. , lupton , r. h. , mckay , t. a. , meiksin , a. , nichol , r. c. , pope , a. c. , schlegel , d. j. , schneider , d. p. , spergel , d. n. , stoughton , c. , strauss , m. a. , szalay , a. s. , tegmark , m. , vogeley , m. s. , weinberg , d. h. , york , d. g. , & zehavi , i. 2009 , , 1741 , s. , aldering , g. , goldhaber , g. , knop , r. a. , nugent , p. , castro , p. g. , deustua , s. , fabbro , s. , goobar , a. , groom , d. e. , hook , i. m. , kim , a. g. , kim , m. y. , lee , j. c. , nunes , n. j. , pain , r. , pennypacker , c. r. , quimby , r. , lidman , c. , ellis , r. s. , irwin , m. , mcmahon , r. g. , ruiz - lapuente , p. , walton , n. , schaefer , b. , boyle , b. j. , filippenko , a. v. , matheson , t. , fruchter , a. s. , panagia , n. , newberg , h. j. m. , couch , w. j. , & the supernova cosmology project . 1999 , , 517 , 565 , f. , ade , p. a. r. , bock , j. j. , bond , j. r. , borrill , j. , boscaleri , a. , cabella , p. , contaldi , c. r. , crill , b. p. , de bernardis , p. , de gasperis , g. , de oliveira - costa , a. , de troia , g. , di stefano , g. , hivon , e. , jaffe , a. h. , kisner , t. s. , jones , w. c. , lange , a. e. , masi , s. , mauskopf , p. d. , mactavish , c. j. , melchiorri , a. , montroy , t. e. , natoli , p. , netterfield , c. b. , pascale , e. , pogosyan , d. , polenta , g. , prunet , s. , ricciardi , s. , romeo , g. , ruhl , j. e. , santini , p. , tegmark , m. , veneziani , m. , & vittorio , n. 2006 , , 647 , 833 , t. , benson , b. a. , ade , p. a. r. , aird , k. a. , bleem , l. e. , carlstrom , j. e. , chang , c. l. , cho , h. , crawford , t. m. , crites , a. t. , de haan , t. , dobbs , m. a. , george , e. m. , hall , n. r. , halverson , n. w. , holder , g. p. , holzapfel , w. l. , hrubes , j. d. , joy , m. , keisler , r. , knox , l. , lee , a. t. , leitch , e. m. , lueker , m. , marrone , d. , mcmahon , j. j. , mehl , j. , meyer , s. s. , mohr , j. j. , montroy , t. e. , padin , s. , pryke , c. , reichardt , c. l. , ruhl , j. e. , schaffer , k. k. , shaw , l. , shirokoff , e. , spieler , h. g. , stalder , b. , staniszewski , z. , stark , a. a. , vanderlinde , k. , vieira , j. d. , williamson , r. , & zahn , o. 2010 , , 716 , 1118 , c. l. , ade , p. a. r. , bock , j. j. , bond , j. r. , brevik , j. a. , contaldi , c. r. , daub , m. d. , dempsey , j. t. , goldstein , j. h. , holzapfel , w. l. , kuo , c. l. , lange , a. e. , lueker , m. , newcomb , m. , peterson , j. b. , ruhl , j. , runyan , m. c. , & staniszewski , z. 2009 , , 694 , 1200 , b. a. , percival , w. j. , eisenstein , d. j. , verde , l. , spergel , d. n. , skibba , r. a. , bahcall , n. a. , budavari , t. , frieman , j. a. , fukugita , m. , gott , j. r. , gunn , j. e. , ivezi , . , knapp , g. r. , kron , r. g. , lupton , r. h. , mckay , t. a. , meiksin , a. , nichol , r. c. , pope , a. c. , schlegel , d. j. , schneider , d. p. , stoughton , c. , strauss , m. a. , szalay , a. s. , tegmark , m. , vogeley , m. s. , weinberg , d. h. , york , d. g. , & zehavi , i. 2010 , , 404 , 60 , a. g. , filippenko , a. v. , challis , p. , clocchiatti , a. , diercks , a. , garnavich , p. m. , gilliland , r. l. , hogan , c. j. , jha , s. , kirshner , r. p. , leibundgut , b. , phillips , m. m. , reiss , d. , schmidt , b. p. , schommer , r. a. , smith , r. c. , spyromilio , j. , stubbs , c. , suntzeff , n. b. , & tonry , j. 1998 , , 116 , 1009 , e. , wechsler , r. h. , rykoff , e. s. , annis , j. t. , becker , m. r. , evrard , a. e. , frieman , j. a. , hansen , s. m. , hao , j. , johnston , d. e. , koester , b. p. , mckay , t. a. , sheldon , e. s. , & weinberg , d. h. 2010 , , 708 , 645 , t. , hartlap , j. , joachimi , b. , kilbinger , m. , simon , p. , benabed , k. , brada , m. , eifler , t. , erben , t. , fassnacht , c. d. , high , f. w. , hilbert , s. , hildebrandt , h. , hoekstra , h. , kuijken , k. , marshall , p. , mellier , y. , morganson , e. , schneider , p. , semboloni , e. , van waerbeke , l. , & velander , m. 2009 , arxiv e - prints , j. l. , achermann , c. , bond , j. r. , bronfman , l. , bustos , r. , contaldi , c. r. , dickinson , c. , ferreira , p. g. , jones , m. e. , lewis , a. m. , mason , b. s. , may , j. , myers , s. t. , oyarce , n. , padin , s. , pearson , t. j. , pospieszalski , m. , readhead , a. c. s. , reeves , r. , taylor , a. c. , & torres , s. 2007 , , 660 , 976 , d. n. , bean , r. , dor , o. , nolta , m. r. , bennett , c. l. , dunkley , j. , hinshaw , g. , jarosik , n. , komatsu , e. , page , l. , peiris , h. v. , verde , l. , halpern , m. , hill , r. s. , kogut , a. , limon , m. , meyer , s. s. , odegard , n. , tucker , g. s. , weiland , j. l. , wollack , e. , & wright , e. l. 2007 , , 170 , 377 , d. n. , verde , l. , peiris , h. v. , komatsu , e. , nolta , m. r. , bennett , c. l. , halpern , m. , hinshaw , g. , jarosik , n. , kogut , a. , limon , m. , meyer , s. s. , page , l. , tucker , g. s. , weiland , j. l. , wollack , e. , & wright , e. l. 2003 , , 148 , 175 , z. , ade , p. a. r. , aird , k. a. , benson , b. a. , bleem , l. e. , carlstrom , j. e. , chang , c. l. , cho , h. , crawford , t. m. , crites , a. t. , de haan , t. , dobbs , m. a. , halverson , n. w. , holder , g. p. , holzapfel , w. l. , hrubes , j. d. , joy , m. , keisler , r. , lanting , t. m. , lee , a. t. , leitch , e. m. , loehr , a. , lueker , m. , mcmahon , j. j. , mehl , j. , meyer , s. s. , mohr , j. j. , montroy , t. e. , ngeow , c. , padin , s. , plagge , t. , pryke , c. , reichardt , c. l. , ruhl , j. e. , schaffer , k. k. , shaw , l. , shirokoff , e. , spieler , h. g. , stalder , b. , stark , a. a. , vanderlinde , k. , vieira , j. d. , zahn , o. , & zenteno , a. 2009 , , 701 , 32 , y. d. , ade , p. a. r. , barkats , d. , battle , j. o. , bierman , e. m. , bock , j. j. , chiang , h. c. , dowell , c. d. , duband , l. , hivon , e. f. , holzapfel , w. l. , hristov , v. v. , jones , w. c. , keating , b. g. , kovac , j. m. , kuo , c. l. , lange , a. e. , leitch , e. m. , mason , p. v. , matsumura , t. , nguyen , h. t. , ponthieu , n. , pryke , c. , richter , s. , rocha , g. , & yoon , k. w. 2010 , , 711 , 1141 , a. , burenin , r. , forman , w. r. , jones , c. , hornstrup , a. , murray , s. s. , & quintana , h. 2007 , in heating versus cooling in galaxies and clusters of galaxies , ed . h. bhringer , g. w. pratt , a. finoguenov , & p. schuecker , 48+ , e. y. s. , ade , p. , bock , j. , bowden , m. , brown , m. l. , cahill , g. , castro , p. g. , church , s. , culverhouse , t. , friedman , r. b. , ganga , k. , gear , w. k. , gupta , s. , hinderks , j. , kovac , j. , lange , a. e. , leitch , e. , melhuish , s. j. , memari , y. , murphy , j. a. , orlando , a. , piccirillo , l. , pryke , c. , rajguru , n. , rusholme , b. , schwarz , r. , osullivan , c. , taylor , a. n. , thompson , k. l. , turner , a. h. , & zemcov , m. 2009 , physical review letters , 102 , 161302 | the combination of 7-year data from _ wmap _ and improved astrophysical data rigorously tests the standard cosmological model and places new constraints on its basic parameters and extensions . by combining the _ wmap _ data with the latest distance measurements from the baryon acoustic oscillations ( bao ) in the distribution of galaxies @xcite and the hubble constant ( @xmath0 ) measurement @xcite
, we determine the parameters of the simplest 6-parameter @xmath1cdm model .
the power - law index of the primordial power spectrum is @xmath2 ( 68% cl ) for this data combination , a measurement that excludes the harrison - zeldovich - peebles spectrum by 99.5% cl .
the other parameters , including those beyond the minimal set , are also consistent with , and improved from , the 5-year results .
we find no convincing deviations from the minimal model .
the 7-year temperature power spectrum gives a better determination of the third acoustic peak , which results in a better determination of the redshift of the matter - radiation equality epoch .
notable examples of improved parameters are the total mass of neutrinos , @xmath3 , and the effective number of neutrino species , @xmath4 ( 68% cl ) , which benefit from better determinations of the third peak and @xmath0 .
the limit on a constant dark energy equation of state parameter from _
wmap_+bao+@xmath0 , without high - redshift type ia supernovae , is @xmath5 ( 68% cl ) .
we detect the effect of primordial helium on the temperature power spectrum and provide a new test of big bang nucleosynthesis by measuring @xmath6 ( 68% cl ) .
we detect , and show on the map for the first time , the tangential and radial polarization patterns around hot and cold spots of temperature fluctuations , an important test of physical processes at @xmath7 and the dominance of adiabatic scalar fluctuations .
the 7-year polarization data have significantly improved : we now detect the temperature-@xmath8-mode polarization cross power spectrum at 21@xmath9 , compared to 13@xmath9 from the 5-year data . with the 7-year temperature-@xmath10-mode
cross power spectrum , the limit on a rotation of the polarization plane due to potential parity - violating effects has improved by 38% to @xmath11 ( 68% cl ) .
we report significant detections of the sunyaev - zeldovich ( sz ) effect at the locations of known clusters of galaxies .
the measured sz signal agrees well with the expected signal from the x - ray data on a cluster - by - cluster basis .
however , it is a factor of 0.5 to 0.7 times the predictions from `` universal profile '' of arnaud et al . , analytical models , and hydrodynamical simulations .
we find , for the first time in the sz effect , a significant difference between the cooling - flow and non - cooling - flow clusters ( or relaxed and non - relaxed clusters ) , which can explain some of the discrepancy .
this lower amplitude is consistent with the lower - than - theoretically - expected sz power spectrum recently measured by the south pole telescope collaboration . |
the observable universe is close to a flat friedmann model , the so - called einstein - de sitter universe , in which the energy density , @xmath3 takes the critical value , @xmath4 , and the homogeneous spatial surfaces are euclidean .
all astronomical evidence shows that we are quite close to this state of flatness , although a value of @xmath5 in the vicinity of @xmath6 is preferred by several observations .
it is therefore disquieting to notice that the flat friedmann model containing dust or blackbody radiation is unstable as time increases .
small deviations from the exact @xmath7 model grow quickly in time , typically like @xmath8 , where @xmath9 is the expansion factor of the universe .
the observed @xmath10 state therefore requires extreme fine tuning of the cosmological initial conditions , assumed to be set at planck epoch , because @xmath11 has increased by a factor of order @xmath12 from that epoch to the present .
this is the _
flatness problem_. if the universe is slightly open at planck time , within a few planck times it would become totally curvature dominated .
if it is initially slightly closed , it would quickly collapse to planck density again . explaining its current state requires an extraordinarily close proximity to perfect flatness initially , or some sequence of events which can subsequently reverse expectations and render the flat solution asymptotically stable .
particle physics theories naturally contain self - interacting scalar matter fields which violate the strong energy condition ( so the density and pressure , @xmath13 , obey @xmath14 ) .
these can make the flat solution asymptotically stable with increasing time , allowing the asymptotic state to naturally be close to flatness .
cosmological histories in which a brief period of expansion is dominated by a matter field or other effective stress which violate the strong energy condition , and so exhibit gravitational repulsion , are called `` inflationary '' .
solutions to the flatness problem have been proposed in the context of inflationary scenarios @xcite , pre - big - bang models @xcite , and varying speed of light cosmologies @xcite . in all of these theories , @xmath7 becomes an asymptotically stable attractor for the expanding universe .
the observed @xmath15 state results then from a temporary period of calculable physical processes , rather than from highly tuned initial conditions . in such scenarios
the price to be paid is that @xmath5 should be very close to unity , @xmath16 @xmath17 , if one is not to invoke an unmotivated fine tuning of the initial conditions again .
if we take the trend of the observational data seriously , then explaining a current value of @xmath5 of , say , @xmath18 is yet another challenge .
we call it the _ quasi - flatness problem_. solutions to this problem have been proposed in the context of open inflationary models @xcite . in these one
has to come to grips with some degree of fine tuning .
the anthropic principle @xcite is usually invoked for this purpose @xcite but considerable uncertainties exist in the range of predictions that emerge and there does not appear to be scope for a very precise prediction of @xmath5 in , say , the range @xmath19 undoubtedly , it would be better if one could find mechanisms which would produce a definite @xmath5 of order one , but different from 1 , as an attractor . in a recent letter @xcite
, we displayed one theory in which this possible .
here we present further solutions in support of such models .
we explore analytical and numerical solutions to brans - dicke ( bd ) cosmologies with a varying speed of light ( vsl ) .
these generalise our earlier investigations of this theory @xcite .
we show that if the speed of light evolves as @xmath20 with @xmath21 there is a late - time attractor at @xmath22 .
hence , these cosmologies can solve the quasi - flatness problem .
this work expands considerably the set of solutions presented in ref .
@xcite .
we note that the existence of the dimensionless fine structure constants allows these varying-@xmath23 theories to be transformed , by a change of units , into theories with constant @xmath24 but with a varying electron change , @xmath25 , or dielectric constant of the vacuum .
this process is described in detail in ref .
@xcite where a particular theory is derived from an action principle .
a different varying-@xmath25 theory has been formulated by bekenstein @xcite but it is explicitly constructed to produce no changes in cosmological evolution .
a study of this theory will be given elsewhere @xcite .
there also exist analogues of the flatness and quasi - flatness problems with regard to the cosmological constant .
the _ lambda problem _ is to understand why the cosmological constant term , @xmath26 in the friedmann equation does not overwhelmingly dominate the density and curvature terms today ( quantum gravity theories suggest that it ought to be about @xmath27 times bigger than observations permit @xcite ) .
the _ quasi - lambda problem _ is to understand how it could be that this contribution to the friedmann equation could be non - zero and of the same order as the density or curvature terms ( as some recent supernovae observations suggest @xcite ) . in section [ eqns ]
we write down the evolution equations in these theories .
the varying @xmath28 aspect of the theory will be accommodated in the standard way by means of a brans - dicke theory of gravitation .
this theory is adapted in a well - defined way to incorporate varying @xmath29 we write equations in both the jordan and einstein s frames . in section [ pert ]
we study solutions to the flatness problem in the perturbative regime ( @xmath30 ) , when both @xmath23 and @xmath28 may change . in section [ gcons ]
we present non - perturbative solutions when @xmath28 is constant , and in section [ gvar ] when @xmath28 is varying . in section [ asym ]
we derive simple conditions on any power - law variation of @xmath23 with scale factor if the flatness problem is to be solved .
these conclusions are reinforced by some exact solutions in section [ exact ] .
in section [ qflat ] we show that the quasi - flatness problem can be naturally solved in a class of varying-@xmath23 cosmologies and then we discuss the solution of the quasi - lambda problem by such cosmologies in section [ qlamb ] .
we conclude with a summary of our results , highlighting the cases in which we can claim to have solved the quasi - flatness and quasi - lambda problems .
in bd varying speed of light ( vsl ) theories the friedmann equations are @xcite : @xmath31-\omega { \left ( \frac{\dot \phi } \phi \right ) } ^2-{\frac { \ddot \phi } \phi } \label{fr1 } \\ { \left ( \frac{\dot a}a\right ) } ^2 & = & { \frac{8\pi \rho } { 3\phi } } -{\frac{kc^2 } { a^2}}-{\frac{\dot \phi a}{\phi a}}+{\frac \omega 6}{\left ( \frac{\dot \phi } \phi \right ) } ^2 \label{fr2}\end{aligned}\ ] ] where the speed of light , @xmath24 is now an arbitrary function of time , @xmath32 is the curvature constant , and @xmath33 is the constant bd parameter .
the wave equation for the bd scalar field @xmath34 is @xmath35 from these three equations we can obtain the generalised conservation equation . since after we impose the equation of state , the varying @xmath23 term only appears in eq .
( [ fr2 ] ) , the only new contribution to @xmath36 is from the @xmath37 term .
hence , together , these imply the non - conservation equation : @xmath38 in the radiation - dominated epoch ( @xmath39 ) the general solution for @xmath40 is @xmath41 if the integration constant @xmath42 the usual solutions for vsl in general relativity with varying @xmath23 follow illustrating the fact that for a radiation source any solution of general relativity is a particular solution of bd theory with constant @xmath43 . in the next sections we explore the solutions to the flatness problem when @xmath44 .
we will also consider the radiation to matter transition in these theories .
equations ( [ fr1])-([dotg ] ) apply in the so - called jordan frame
. it will be useful to introduce the einstein frame , by means of the transformations , @xmath45 which are performed at constant @xmath23 .
these may be regarded as merely mathematical transformations of variables .
the friedmann equations in the new frame are @xmath46-{\frac{\sigma ^{\prime } { } ^2}3 } \\ { \left ( \frac{{\hat a}^{\prime } } { { \hat a}}\right ) } ^2+{\frac{kc^2}{{\hat a } ^2 } } & = & { \frac{8\pi g{\hat \rho } } 3}+{\frac{\sigma ^{\prime } { } ^2}6}\end{aligned}\ ] ] where @xmath47 .
the transformed scalar field equation for @xmath48 is @xmath49 these are just the standard friedmann equations with constant @xmath28 and a scalar field added to the normal matter .
the scalar field behaves like a stiff perfect fluid with equation of state @xmath50 in the einstein frame , if @xmath51 , the total stress - energy tensor is conserved , but the scalar field and normal matter exchange energy according to : @xmath52 if @xmath53 one has instead @xmath54 all equations derived for standard vsl theory , with constant @xmath28 , are valid in the einstein frame .
however , one should always remember to add to normal matter the scalar field energy and pressure ( so the total density and pressure are given by @xmath55 and @xmath56 respectively ) .
we first study solutions to the flatness problem when there are small deviations from flatness .
let us define the critical density , @xmath4 , in a b - d universe by means of the equation : @xmath57 that is , eqn .
[ fr2 ] with @xmath58 . in the einstein frame
the critical density in normal matter is @xmath59 in terms of total energy density @xmath55 , the critical energy density in the einstein frame is : @xmath60 accordingly , we may define a _ relative flatness parameter : _ @xmath61 as shown in the previous section , the usual equations for vsl theory should apply to this quantity .
therefore , one has @xmath62 with the equation of state given by @xmath63 with @xmath64 constant .
but , since @xmath65 we see that the natural quantity to quantify deviations from flatness in the jordan frame is not @xmath66 but an adaptation of it , @xmath67 which satisfies the equation @xmath68 if @xmath69 this can be integrated to give @xmath70 hence , we have the solution @xmath71 if @xmath72 varies then the second term on the right always works against solving the flatness problem . hence solving the flatness problem in bd
requires that @xmath43 decreases in the early universe and that the first term on the right of eq .
( [ perteps ] ) dominates the second . in a radiation - dominated universe ( @xmath73 ) : @xmath74 where @xmath75 was defined by eqn.([dotg ] ) .
if @xmath75 is small and @xmath76 one still solves the flatness problem when @xmath77 . notice that since @xmath78 at late times , the second term always eventually becomes negligible
in order to explore non - perturbative solutions to the flatness problem we solved eqns .
( [ fr1])-([dotg ] ) numerically .
first we consider the case where @xmath80 .
exact solutions exist in this case for matter and radiation - dominated universes . setting @xmath34 in eqn .
( [ eps ] ) leads to @xmath81 which , assuming @xmath82 becomes @xmath83 the general structure of the attractors can be inferred from this equation . however , it can also be integrated to give @xmath84 with @xmath85 , and where the integration constant @xmath86 is chosen so as to enforce the specified initial conditions .
our numerical code matched this analytical result to very high accuracy . in fig .
[ fig1 ] we plot numerical results for a radiation - dominated universe .
we comment on four possible situations , illustrated in figs .
[ fig1]-[fig4 ] .
if @xmath23 is a constant ( @xmath87 ) then the flat universe ( @xmath88 ) is unstable ( fig .
[ fig1 ] ) .
there are two attractors at @xmath89 and @xmath90 , showing that the universe tends to become curvature dominated . for slightly closed universes
this means evolution towards a big crunch singularity . for slightly open universes
it means evolution towards a milne universe ( @xmath89 ) which in the case of constant @xmath28 is simply an expanding empty spacetime with @xmath91 .
if the speed of light were to increase with @xmath9 ( @xmath92 with @xmath93 ) the situation is the same , but with a stronger repulsion from @xmath88 .
if @xmath92 with @xmath94 , then , as pointed out in @xcite , the flatness problem is solved if @xmath95 is not too far from zero initially .
in fact , @xmath88 is now an attractor ( see fig . [ fig2 ] ) .
the non - perturbative analysis reveals two novelties : there is an unstable node at @xmath96 and @xmath97 is an attractor .
this means that if the universe at any given initial time satisfies @xmath98 it will not evolve towards the flat attractor .
instead , it will evolve towards a big crunch with @xmath99 .
therefore , for closed universes , the flatness problem is solved only if they are not too positively curved .
it is curious to note that the situation is different for open universes .
no matter how far they are from flatness they will always evolve towards the einstein - de sitter model at late times .
this was first pointed out in @xcite , where it was noted that in vsl theories the einstein - de sitter universe is an attractor even if the universe is initially minkowski ( empty ) or de sitter ( dominated by a cosmological constant ) .
the greatest novelty , however , appears for the case in which @xmath100 .
although @xmath88 is now unstable , there is a stable attractor at @xmath96 ( see fig . [ fig3 ] ) .
the system is now repelled from @xmath89 but is attracted to @xmath99 .
hence we have the following scenario : all closed universes evolve towards a big crunch , no matter how close to flatness they are initially .
thus a selection process excludes even slightly closed universes . _ _ in contrast , flat and all open models ( no matter how low their density ) evolve towards an open model with @xmath101 , a value that lies between @xmath102 and @xmath103 , and is typically of order unity .
we therefore have a solution to the quasi - flatness problem , for open , but not for closed universes .
the @xmath104 case is a transition case in between the last two situations just described ( fig .
[ fig4 ] ) . in this case
@xmath88 is a saddle : stable if approached from below but unstable from above . in quasi - flat scenarios the major contribution to the matter density of the universe
is the term resulting from violations of energy conservation .
this can be seen from @xmath105 derived above . taking radiation as an example ( @xmath73 ) ,
if @xmath77 we are pushed towards the attractor @xmath7 and the energy - violation term ( second term ) becomes negligible . in scenarios
in which the quasi - flatness problem is solved the converse happens : the second term dominates .
matter is continuously being created and this provides the dominant contribution to cosmic matter at any given time . matter which was created in the past gets diluted away by expansion very quickly . in effect
, we have a state of permanent `` reheating '' .
this scenario has a passing resemblance with the steady state universe , in which the so called @xmath86 field continuously creates more and more matter .
if we couple a changing @xmath23 say to the lagrangian of the standard particle physics model then all particles are produced , and whether they behave like matter or radiation depends on whether @xmath106 or @xmath107 .
if @xmath85 the energy density in created matter decreases , and the universe cools down as it expands . hence there will necessarily be a radiation to matter transition in these scenarios , just like in the standard big bang theory with constant @xmath108 and @xmath23 .
a complication arises when we attempt to model the evolution through this transition because the structure of the attractors changes . for a radiation - dominated universe
one requires @xmath109 for the open attractor @xmath110 to be achieved . for a matter - dominated universe
the attractors for open universes are given by @xmath111 and are achieved when @xmath112 .
hence , if @xmath113 we have attraction towards a ( quasi - flat ) open universe during the radiation - dominated epoch which is then driven towards flatness in the subsequent matter - dominated epoch . if @xmath114 we have attraction to a ( quasi - flat ) open model in both epochs but the universe will evolve from a trajectory asymptoting to @xmath22 toward one asymptoting to @xmath111 after the matter - radiation equality time .
the solution ( [ soleps ] ) is then no longer valid because @xmath64 is time dependent .
for radiation , @xmath73 , we can solve the field equations by generalising the method introduced for scalar - tensor theories introduced in ref .
define the conformal time @xmath116 so @xmath117 eq .
( [ fr3 ] ) integrates to give @xmath118 with @xmath119 constant .
now change the variables of eq.([fr2 ] ) by introducing @xmath120 and @xmath121 it reduces to @xmath122 for radiation , eq . ( [ den1 ] ) gives @xmath123 if we substitute ( [ y2 ] ) in ( [ y1 ] ) then we have @xmath124\ y^{2(m+1 ) } \label{y5}\ ] ] this can be integrated exactly for appropriate choices of @xmath125 and its qualitative behaviour is easy to understand .
note that if we can solve ( [ y5 ] ) for @xmath126 @xmath127 , then we know @xmath128 and can solve ( [ int ] ) to get @xmath129 hence @xmath130 and finally @xmath131 from ( [ time ] ) since @xmath132 we note that the curvature term in ( [ y5 ] ) changes sign for a special value of @xmath133 @xmath134 examining ( [ y5 ] ) we see that the @xmath58 solution appears to be approached if @xmath135 , ie @xmath136 when @xmath137 there is a simple special case which allows an exact solution .
( [ y5 ] ) is now @xmath138 where @xmath139 we integrate ( [ y6 ] ) to get @xmath140 hence , @xmath141 ^{a/\sqrt{\alpha + \gamma } } \ ] ] and @xmath142 ^{a/\sqrt{\alpha + \gamma } } \ ] ] we see that if @xmath143 then , as @xmath144 we have @xmath145 unless the denominator blows up .
if @xmath146 then we get @xmath147 and @xmath148 \exp \{-\frac{4a}\lambda \tan ^{-1}\left ( \frac{\beta ( \tau + \tau _ 0 ) } \lambda \right ) \}\ ] ] where @xmath149 looking back at the defining equation ( [ y5 ] ) , we see that for this @xmath137 case , @xmath150 and so @xmath151 and for large @xmath119 we have @xmath143 when @xmath152 .
a numerical evolution of the equations with a varying @xmath43 reveals that for reasonable initial values of @xmath43 the structure of attractors is not affected .
however , the speed at which attractors are reached may be increased if @xmath43 is allowed to change . as an example we shall consider the case in which one starts from a milne universe ( @xmath89 ) . in @xcite
it was argued that this is the natural initial condition to consider in the context of vsl cosmologies .
let us then consider the case @xmath73 , and integrate eqns .
( [ fr1 ] ) , ( [ fr2 ] ) , and ( [ fr3 ] ) .
we consider solutions of the form ( [ dotg ] ) with various integration constants @xmath75 . in figures
[ figbd ] and [ figbd1 ] we show the evolution of @xmath153 and @xmath43 from such a state , with @xmath154 and @xmath155 . in the first case
we see that we still have a flat attractor , which is reached much more rapidly if @xmath40 is increasing or decreasing . in the latter case
we have a quasi - flat attractor , with @xmath156 , whether or not @xmath43 is allowed to change .
the attractor is achieved faster with a changing @xmath43 , especially a decreasing one .
exact solutions are only possible for particular equations of state and @xmath23-variation laws but it is possible to understand the asymptotic behaviour in general .
the vsl theory brans - dicke solutions are given for eqns .
( [ fr1])-([fr3 ] ) together with ( [ noncons ] ) .
we assume a perfect - fluid equation of state given by ( [ gamma ] ) .
in the flat case ( @xmath157 ) the equations reduce to those of standard bd flat universes with constant @xmath23 . at late times , when @xmath58 , the general bd solutions for all @xmath158 approach the particular ( matter - dominated or machian ) power - law solutions @xcite @xmath159 where @xmath160 @xmath161 hence , we see that @xmath162 note that @xmath163 for radiation;.these exact power - law solutions reduce the special @xmath164 general relativity solution in the radiation case
. we shall therefore look at the stability of these @xmath58 asymptotes when we turn on the @xmath165 terms in eqns .
( [ fr1])-([fr3 ] ) and ( [ noncons ] ) with @xmath166 included . if we substitute ( [ a ] ) and ( [ d ] ) in ( [ noncons ] ) with @xmath76 then we have @xmath167 integrating , we get @xmath168 if @xmath169 . thus , to solve the flatness problem we will need the @xmath170 term to dominate the @xmath32 term on the right - hand side of eq .
( [ asy ] ) at large @xmath171 that is , since @xmath172 for expanding universes , asymptotic approach to flatness requires @xmath173 using ( [ a ] ) , this gives the condition @xmath174}.\ ] ] for radiation and dust universes this condition for solving the flatness problem reduces to : @xmath175 where ( [ co2 ] ) approaches the general relativity value for @xmath176 as expected and ( [ co1 ] ) agrees with the radiation case studied above .
the exact solutions found for the vsl theory with @xmath154 in @xcite can be generalized to the bd case if one assumes that the variation of the speed of light is governed by a relation of the form @xmath177 this reduces to the previously studied ( non - bd ) case when @xmath43 is constant . in the einstein frame one
then has @xmath178 and in the radiation - dominated epoch the first of equations ( [ conseins ] ) becomes @xmath179 hence , for radiation , one has the integral @xmath180 with @xmath170 constant . since @xmath181 in the jordan frame
, one has @xmath182 this solution can be verified directly from eq .
( [ noncons ] ) , although it would have been difficult to guess it without a foray into the einstein frame .
conditions for solving the flatness problem can now be derived by inspection of the friedmann - like equation : @xmath183 since @xmath43 approaches an asymptotic value we still require that @xmath77 for the term in @xmath184 to dominate the curvature ( @xmath32 ) term at late times .
if @xmath43 is decreasing we do not need the speed of light to decrease in time so fast in order to solve the flatness problem . since @xmath185
we see that while @xmath186 is non - negligible we have @xmath187 .
weaker gravity in the early universe therefore assists a varying speed of light in solving the flatness problem introduce the variable @xmath188 and assume that @xmath23 varies as @xmath189 hence , ( [ cons2 ] ) becomes @xmath190 integrating , we have @xmath191 if @xmath192 .
this solution can be used to study the evolution for general @xmath193
we want to discover if it is ever natural to have evolution which asymptotes to a state of expansion with a _ non - critical _ density ( for example , say , @xmath194 as some observations have implied ) . for simplicity
we consider first the solutions with constant @xmath195 the conservation equation ( [ noncons ] ) with constant @xmath43 reduces to @xmath196 which integrates to give @xmath197 with @xmath170 constant , so substituting in ( [ fr2 ] ) we have @xmath198 > from this we can easily determine the attractors at large @xmath9 . specialising to the radiation case , we have @xmath199 now , the density parameter @xmath200 is defined by : @xmath201 for a quasi - flat open universe the ratio between the two terms on the right - hand side is approximately constant . from the solution ( [ exa ] ) with @xmath73
we have @xmath202 if @xmath77 the energy production term ( second term on the right - hand side of ( [ exa ] ) ) is subdominant at late times , and so this ratio goes to infinite , meaning @xmath7 .
if @xmath21 then the second term in ( [ exa ] ) dominates at late times and therefore the expansion asymptotes to one displaying @xmath203 that is @xmath22 .
so the key feature is that in these scenarios the curvature terms associated with violations of energy conservation dump energy into the universe at the same rate as the curvature term in friedmann equation .
therefore the ratio of the two terms in the friedmann equation stays constant , leading to an open universe with finite @xmath200 value today .
the formulation of the radiation case in terms of the variable @xmath204 allows us to extend the analysis to the varying-@xmath43 case . taking @xmath205 we have @xmath206 and @xmath207 we introduce the density parameter @xmath208 so that @xmath209 .\ ] ] as @xmath210 we see that @xmath211 if @xmath212 ( in section _ [ gvar ] _ we give a full exact solution for the case @xmath137 which falls into this class ) , but if @xmath213 we have approach to a quasi - flat open universe with @xmath214 again , we can have a natural quasi - flat asymptote with @xmath215 consider the behaviour of eq .
( [ y5 ] ) at large @xmath216 and @xmath217 in the case where @xmath218 that is , where we have the quasi - flat attractor @xmath219 as @xmath220 this assumes the constants in the friedmann equation allow sufficient expansion to occur ( so there is no collapse at a finite early time ) .
since @xmath221 we have @xmath222 ^{\frac 2{\omega _ \infty } } \sim \tau ^{\frac 2{\omega _ \infty } } \ ] ] as @xmath220 using ( [ y6 ] ) we get @xmath223\ ] ] so , since @xmath224 , we have @xmath225 as @xmath226 so , we have @xmath227 \sim \tau ^{\frac 1{\omega _ \infty } } \ ] ] thus @xmath228 and @xmath229 as @xmath230 when @xmath231 we have the expected @xmath232 flat radiation asymptote .
let us consider the impact of a varying speed of light on the general relativity case .
similar results will occur in the bd case ( to which it reduces exactly in the case of radiation with constant @xmath43 ) .
if we wish to incorporate a cosmological constant term , @xmath1 , ( which we shall assume to be constant ) then we can define a vacuum stress obeying an equation of state @xmath233 where @xmath234 then , since @xmath28 is constant , and replacing @xmath235 by @xmath236 in ( [ noncons ] ) , we have the generalisation @xmath237 we shall assume that the matter obeys an equation of state of the form ( [ gamma ] ) and the friedmann equation is @xmath238 we also assume that @xmath154 again , so eq . ( [ cons2 ] ) integrates immediately to give @xcite @xmath239 with @xmath170 a positive integration constant . substituting in ( [ fr ] ) we have @xmath240 eq .
( [ frnew ] ) allows us to determine what happens at large @xmath241 if @xmath242 then we see that the flatness and lambda problems are both solved as before .
there are three distinct cases : the @xmath1 term dominates the right - hand side of ( [ frnew ] ) , the curvature term becomes negligible , and @xmath244 so , at large @xmath245we have @xmath246 note that for radiation this case requires @xmath247 while for dust it requires @xmath248 for general fluids it requires @xmath249 the scale factor approaches ( [ lamas ] ) but the curvature term dominates the matter density term
. define @xmath251 and then we have that , at large @xmath9 , @xmath252 thus @xmath253 @xmath254 as @xmath255 for @xmath256 but when @xmath257 @xmath258 in ( [ frnew ] ) , and so @xmath211 . if @xmath259 we note that @xmath260 when @xmath261 and this is just the solution for the quasi - flatness problem found above for general relativity and brans - dicke theory in section [ qflat ] . for @xmath262 when the @xmath1 term dominates at large @xmath9
we see that @xmath263 and we have a solution to the quasi - lambda problem ( _ ie _ the problem of why @xmath264 and @xmath265 are of similar order today ) .
recall that in this case we have @xmath85 , @xmath266 and so the asymptote is again of the form @xmath267
scenarios in which the quasi - flatness problem is solved are considerably more exotic than the vsl solution to the flatness problem . unlike flat scenarios they have a planck epoch , something which may be a problem .
we discuss this issue in the appendix .
these scenarios possess several other unusual features . in standard flatness vsl scenarios ,
the expansion factor in the radiation - dominated phase is still @xmath268 .
standard nucleosynthesis should still be valid unless there are changes to other aspects of relevant strong and weak interaction physics ( which seems likely ) .
however in scenarios which solve quasi - flatness we have @xmath269 , which for a @xmath270 attractor means @xmath271 .
this could easily conflict with the nucleosynthesis constraints .
however it is not enough to state that the expansion factor at nucleosynthesis time is different : the couplings , masses , decay times , etc , going into nucleosynthesis are all different @xcite .
one must rework the whole problem from scratch before ruling out these scenarios on grounds of discordant nucleosynthesis predictions .
structure formation is another concern .
it was shown in @xcite that the comoving density contrast @xmath272 and gauge - invariant velocity @xmath273 are subject to the equations : @xmath274 where @xmath275 is the comoving wave vector of the fluctuations , and @xmath276 is the entropy production rate , @xmath277 the anisotropic stress , and the speed of sound @xmath278 is given by @xmath279 note that the thermodynamical speed of sound is given by @xmath280 . since in standard big bang models evolution
is isentropic : @xmath281 .
when @xmath282 the evolution need not be isentropic .
however , we keep the definition @xmath283 since this is the definition used in perturbative calculations .
one must however remember that the speed of sound given in ( [ cs ] ) is not the usual thermodynamical quantity . with this definition one
has @xmath284 for adiabatic perturbations ; that is , the ratio between pressure and density fluctuations mimics the ratio of its background rate of change .
let us assume a radiation - dominated background . for superhorizon modes ( @xmath285 )
there is a power - law solution , @xmath286 with @xmath287 and @xmath288 .
the general solution takes the form : @xmath289 where @xmath119 and @xmath170 are constants in time . for a constant @xmath23 ( @xmath87 ) this reduces to the usual @xmath290 growing mode , and @xmath291 decaying mode .
if the flatness problem is to be solved , one must have @xmath77 .
if this condition is satisfied there is no growing mode .
this is an expression , in the context of machian bd scenarios , of the link between solving the flatness problem and suppressing density fluctuations .
flatness is imposed in vsl by violations of energy conservation , acting so as to leave the universe is a state with @xmath7 .
this process acts locally , so it also suppresses density fluctuations .
alternatively , we can see that the approach to flatness everywhere means that inhomogeneous variations in the spatial curvature must also die away .
accordingly , we see that in vsl scenarios there is a strong connection between solving the flatness problem , and predicting a perfectly homogeneous universe .
quasi - flat scenarios therefore risk not solving the homogeneity problem . on the other hand there
could some mechanism for amplifying thermal fluctuations to become seeds for the large scale structure of the universe @xcite .
indeed the equations above should provide a transfer function converting the thermal white noise spectrum into a tilted or flat spectrum .
if this were the case , then these theories would predict a link between the spectral tilt and @xmath200 .
finally , one may wonder whether such @xmath166 could ever arise in a dynamical vsl theory .
in work in preparation @xcite we address this issue , with the result that indeed if @xmath292 is a scalar field with a brans dicke type of dynamics : @xmath293 then one has a machian solution with an exponent @xmath294 related to the coupling @xmath33 of the theory . hence there will be a range of coupling values for which the flatness problem is solved , and a range for which the quasi - flatness problem is solved . in the latter case , not only
can we predict an open attractor but also the value of @xmath200 of the universe is related to a coupling constant .
we have performed an extensive analysis solutions to brans - dicke theories , with varying @xmath28 , in which the speed of light is also permitted to vary in time .
we have found cosmological scenarios with novel features . for power - law variations in the velocity of light with the cosmological scale factor we identified the cases where the flatness problem can be solved .
these generalise the conditions found in earlier investigations of this vsl theory . unlike in inflationary universes which solve the flatness problem ,
no unusual matter fields are required in the early universe .
we have also identified the cases in which the quasi - flatness problem can be solved ; that is , where there can be asymptotic approach at late times to an open universe with a density close to that of the critical value .
similarly , we identified those variations of @xmath23 which provided solutions of the lambda problem and the quasi - lambda problem .
the possibility of solving the quasi - flatness and quasi - lambda problems in this way is a genuine novelty of the vsl theory that distinguishes from the standard inflationary universe scenario .
we have also discussed some problems with the vsl scenario , highlighting in particular the matter - radiation transition solutions and the role of the planck epoch in setting initial conditions ( see appendix for further details ) .
we have also examined the behaviour of inhomogeneous perturbations to the homogeneous and isotropic solutions and found the conditions for density perturbations to grow or decay .
our investigations reinforce the conclusions of our earlier investigations of varying-@xmath23 theories without varying @xmath28 : the scope for obtaining cosmological models which share a number of appealing properties , which closely mirror those of the observed universe , suggest that cosmologies with varying @xmath23 should be thoroughly explored .
it is a challenge to find observational predictions which would allow future satellite probes of the microwave background radiation structure to distinguish them from inflationary universe models ( with constant @xmath23 ) .
we hope that this paper will serve as a further stimulus to undertake those investigations and to search out new ways of testing the constancy of the traditional constants of nature @xcite .
jm acknowledges financial support from the royal society and would like to thank a. albrecht and c. santos for helpful comments .
jdb is supported by a pparc senior fellowship .
the machian vsl scenario , in which @xmath154 , introduced by barrow @xcite has significant advantages to the phase transition scenario , in which the speed of light changes suddenly from @xmath295 to @xmath296 preferred by albrecht and magueijo @xcite . in the phase transition scenario one runs into the problem of having to decide when to lay down `` natural initial conditions '' ( that is @xmath95 and @xmath297 of order 1 ) . for a phase transition occurring at time @xmath298 ,
the planck time @xmath299 ( built from the constants as they were before the transition ) is much smaller than @xmath300 . but why should we lay down natural initial conditions just before the phase transition ? if the only scales in the problem are the ones set by the constants as they were before the transition , then natural initial conditions should be set at @xmath301 .
if we are to lay down natural initial conditions at @xmath301 then the universe goes off the attractor well before the phase transition .
a catastrophic phase precedes the phase transition , in which the universe becomes milne ( curvature dominated ) or de sitter ( @xmath1 dominated ) .
albrecht and magueijo noted that this catastrophic phase is not the end of the universe in vsl scenarios .
a varying speed of light would not write off milne or de sitter universes , but would still push them towards an einstein - de sitter universe
. the only universes which would be selected out in this process are the ones with positive curvature , which would end in a big crunch , well before the phase transition .
the machian scenario does not have this problem , if we are content with solving the flatness but not the quasi - flatness problem .
if we consider the radiation dominated phase , @xmath154 with @xmath77 , and @xmath268 , then we have @xmath302 .
hence there is no planck time in these scenarios : as @xmath303 , one has @xmath304 . as we go back in time , the universe becomes hotter and hotter , but the planck temperature also increases , and is never achieved at any time .
the idea of setting natural initial conditions at planck time does not make sense in these scenarios .
we have a universe constantly pushed towards an attractor , which is flat , and has zero cosmological constant .
s. perlmutter et al , ap .
j. * 483 * , 565 ( 1997 ) ; s. perlmutter et al ( the supernova cosmology project ) , nature * 391 , * 51 ( 1998 ) ; garnavich , p.m et al , ap.j .
letters * 493 * , l53 ( 1998 ) ; schmidt , b.p .
1998 ap.j . *
507 * , 46 ; riess , a.g .
et al , aj 116 , 1009 ( 1998 ) m.j .
drinkwater , j.k .
webb , j.d .
barrow , and v.v .
flambaum , mon . not .
* 298 , 457 ( 1998 ) ; * j.k .
webb , v.v .
flambaum , c.w .
churchill , m.j .
drinkwater , and j.d .
barrow , phys .
( 1999 in press ) . | we define the flatness and quasi - flatness problems in cosmological models .
we seek solutions to both problems in homogeneous and isotropic brans - dicke cosmologies with varying speed of light .
we formulate this theory and find perturbative , non - perturbative , and asymptotic solutions using both numerical and analytical methods . for a particular range of variations of the speed of light
the flatness problem can be solved . under other conditions
there exists a late - time attractor with a constant value of @xmath0that is smaller than , but of order , unity .
thus these theories may solve the quasi - flatness problem , a considerably more challenging problem than the flatness problem .
we also discuss the related @xmath1 and quasi-@xmath2 problem in these theories .
we conclude with an appraisal of the difficulties these theories may face . |
the layered crystalline structure of the cuprate high - temperature superconductors leads to existence of two types of vortices in these materials , pancake - vortex ( pv ) stacks @xcite induced by the magnetic field component perpendicular to the layers and josephson vortices ( jvs ) @xcite created by the magnetic field component parallel to the layers .
repulsive interaction between the vortices of each type results in formation of regular vortex lattices . in particular , the magnetic field applied along the layers generates triangular lattice of the jvs stretched along the layer direction .
the anisotropy factor @xmath3 sets the important field scale , @xmath4 , where @xmath5 is the interlayer periodicity .
when the magnetic field exceeds @xmath6 the josephson vortices homogeneously fill all layers forming a dense lattice @xcite . in highly anisotropic materials , like bi@xmath0sr@xmath0cacu@xmath0o@xmath2 ( bscco )
this field scale is rather moderate @xmath7 0.5 tesla . in bscco , due to a very weak josephson interlayer coupling
, two types of vortices can coexist in the tilted magnetic field @xcite .
the static attractive interaction between jvs and pv stacks @xcite leads to many exotic vortex states , such as mixed chain - lattice state @xcite and pancake - chains state@xcite , and has been used for visualization of jv stacks @xcite , see recent review @xcite .
dynamic properties of the crossing - lattices state have been studied in much less details .
a particularly interesting topic is dynamic behavior of the jv lattice .
an external transport current flowing across the layers drives the jv lattice along the layers .
due to relatively weak intrinsic dissipation , the josephson vortices can be accelerated up to very high velocities .
dynamics of the jv lattice in bscco have been extensively studied by several experimental groups ( see , e.g. , refs . ) .
when magnetic field is tilted at small angle with respect to the layers , the c - axis field component generates small concentration of pv stacks .
alternating in - plane supercurrents of static jv lattice force zigzag deformations of the pv stacks @xcite , see fig .
[ fig : jvl - pancstack ] .
it is well known that mobility of jvs is strongly suppressed by a very small amount of pv stacks @xcite . as a consequence , studying the jv lattice transport always requires a very accurate alignment of the magnetic field with the layers . in spite of that common knowledge , jv lattice dynamics in presence of the pv stacks has never been investigated systematically . in the case of strong anisotropy
, the jv lattice can move through static pv stacks . even in this case
the pv stacks will strongly impede motion of the jv lattice .
dynamic behavior of the pv stack naturally follows from its static configuration .
the moving jv lattice forces oscillations of the pv stacks leading to additional dissipation and slowing down the lattice . in this paper
we investigate this effect quantitatively in the case of dense jv lattice .
influence of the pv stacks on motion of an _ isolated _ jv has been considered theoretically in ref .. the paper is organized as follows . in sec .
[ sec : theory ] we present theoretical models describing influence of the pv stacks on motion of the dense jv lattice .
we compute the dynamic friction force generated by pv stacks and study suppression of this force by pv fluctuations .
we also consider influence of the pv fluctuations on the regular jv flux - flow resistivity and influence of columnar defects on the pv - induced damping of the jv lattice . in sec.[sec : experiment ] we present experimental results . studying the flux - flow of the jv lattice for small c - axis magnetic fields , we extracted the pv contribution to the jv damping and studied its dependence on the voltage , temperature , and in - plane field .
we also found that this pv contribution is strongly suppressed by heavy - ion irradiation . in sec .
[ sec : discussion ] we discuss comparison between the experiment and theory and possible applications of the studied effect .
a general description of jv dynamics in presence of the pv stacks is rather complicated .
we consider first the simplest situation , in which influence of the pancake vortices on motion of the jv lattice can be traced in a clearest way .
we assume that a strong magnetic field , @xmath8 tesla , is applied in the @xmath9-@xmath10 plane at a very small angle with respect to the layers .
such a magnetic field generates the dense jv lattice and a dilute array of the pv stacks .
we neglect both pinning and thermal fluctuations of the pancake vortices .
we start with derivation of the interaction between the jv lattice and deformed pv stack via the interlayer josephson energy , @xmath11 where @xmath12 is the phase difference between the layers @xmath13 and @xmath14 .
the static dense jv lattice is characterized by the phase difference , @xmath15 with @xmath16 and @xmath17 describing the jv lattice displacement . the deformed pv stack with the coordinates @xmath18 generates the additional phase difference @xmath19 and modifies the josephson energy .
in addition to interactions mediated by interlayer josephson coupling , pvs are also aligned by the magnetic interaction @xcite . due to its long - range nature
, this interaction can be well approximated by a simple parabolic potential ( see , e.g. , ref . ) with the magnetic - spring constant , @xmath20 where @xmath21 and @xmath22 is the typical wandering distance .
therefore in a homogeneous superconductor the total energy change per one layer due to pv stack deformation is given by the sum of the josephson and magnetic energies @xmath23 , \label{josentotal}\ ] ] where @xmath24 is the total number of layers and the function @xmath25 in the lowest order with respect to the josephson coupling energy , @xmath26 , is given by @xmath27 \right.\nonumber\\ & \left.-\cos\left [ k_h(y - y_{j})+\theta _ { p}(\mathbf{r},\mathbf{u}_{2},\mathbf{u}_{1})\right ] \right\ } \label{cdef}\\ = \frac{i(k_h\mathbf{u}_{12})}{k_h^{2}}&\cos\left [ k_h\left ( y_{j}-\frac { u_{1,y}+u_{2,y}}{2}\right ) \right ] \nonumber\end{aligned}\ ] ] with @xmath28 and @xmath29 \label{idef}\end{aligned}\ ] ] with @xmath30 .
in the regime @xmath31 , this dimensionless function can be computed analytically up to the third order with respect to the reduced variable @xmath32 , @xmath33 \label{iexpan}\ ] ] with @xmath34and @xmath35 ( @xmath36 is the euler constant ) .
the linear in @xmath37 term in @xmath38 gives the linear - displacement contribution to the josephson energy , @xmath39 $ ] .
this term describes forces acting on the straight pv stack from the alternating in - plane currents induced by the jv lattice , @xmath40 $ ] ,
where @xmath41 , @xmath42 is the reduced magnetic field and @xmath43 is the josephson current density . due to this term
the ground state corresponds to the alternating pv deformations along the direction of the in - plane field @xcite , @xmath44 with @xmath45 $ ] , see fig .
[ fig : jvl - pancstack ] . for such deformations ,
the quadratic term in the josephson energy cancels out and does not influence the deformation amplitude . the assumed condition
@xmath46 is satisfied if @xmath47 .
lets consider now dynamic behavior .
a transport current applied across the layers drives the jv lattice along the layers .
we consider the lattice slowly moving with constant velocity through the pv stacks along the @xmath48 axis , @xmath49 .
such motion generates the electric field , @xmath50 .
the lattice velocity is assumed to be much smaller than the swihart velocity so that the lattice preserves its static structure .
the homogeneously moving jv lattice forces pvs to oscillate with the josephson frequency , @xmath51 , with @xmath52 and @xmath53 . for a pv stack located near @xmath54 , the alternating deformation , @xmath55 ,
is described by a simple oscillator equation @xmath56 where @xmath57 and @xmath58 is the pancake viscosity coefficient .
solution of this equation is given by @xmath59 .
\label{solution}\ ] ] the frequency response of the stacks is determined by the relaxation frequency , @xmath60 .
the oscillation amplitude drops at @xmath61 .
the average friction force per pv acting on the moving jv lattice is given by @xmath62 \sin\left [ \omega_e t\!-\!k_h\frac{u_{n , y}\!+\!u_{n+1,y}}{2}\right ] \right\rangle \label{frictforce}\ ] ] where @xmath63 notates time averaging .
in the regime when only small alternating deformations in the @xmath9 direction are present , this gives @xmath64 , and , using the solution ( [ solution ] ) , we obtain @xmath65 the average velocity of the jv lattice is connected with the applied current , @xmath66 , by the force - balance condition @xmath67 where @xmath68 is the bare flux - flow friction coefficient of the jv lattice and @xmath69 is the concentration of pvs , and the third term is the driving force from the transport current .
this corresponds to the following current - voltage dependence @xmath70 where @xmath71 is the bare flux - flow conductivity and @xmath72 is the current enhancement due to the pv - induced damping of jv motion , which we will call _ excess pancake - vortex current _ , with @xmath73 and @xmath74k_{p}/\eta_{p}$ ] is the voltage drop per junction corresponding to the relaxation frequency .
the electric - field dependence of the conductivity @xmath75 resembles a well - known drude frequency dependence . introducing the pancake flux - flow conductivity @xmath76
, we can express the conductivity and voltage scales , @xmath77 and @xmath78 , via experimentally - accessible parameters @xmath79 with @xmath80 .
note that @xmath77 scales with the field components as @xmath81 .
the key feature of the excess pv current , @xmath82 , is that it depends nonmonotonically on the electric field .
the maximum current due to the pancake vortices can be estimated as @xmath83 .
as @xmath84 , the total i - v dependence ( [ ivlowt ] ) becomes nonmonotonic at sufficiently large @xmath85 . in fixed - current experiments this leads to voltage jumps .
( [ ivlowt ] ) also determines the angular dependence of voltage at fixed current frequently measured experimentally . introducing the tilt angle of field , @xmath86 , we can rewrite eqs.([ivlowt ] ) and ( [ panccurr ] ) in the form @xmath87 with @xmath88 $ ] .
this equation explicitly determines the dependence @xmath89 at fixed @xmath90 .
in particular , in the case @xmath91 , the maximum and minimum of the dependence @xmath92 are given by @xmath93 at these angles the voltage will have jumps .
as we see , in a simple model both angles are proportional to the in - plane field .
the angular dependence of the electric field following from eq.([angle_ez ] ) at @xmath94 is illustrated in fig .
[ fig : ez_thetatheor ] . ) at @xmath95 .
voltage jumps are expected at the angles @xmath96 and @xmath97 in real experimental conditions the result may be strongly influenced by pinning and thermal fluctuations .
let us consider influence of thermal fluctuations . at a finite temperature , in addition to the regular zigzag deformations
, the pv stack has random fluctuating displacement @xmath98 .
these pancake fluctuations influence the damping of the jv lattice in several different ways .
the most direct channel is that , due to random displacement in the @xmath48 direction ( perpendicular to the field ) , exact time periodicity of the force from the jv lattice acting on pv is perturbed .
this leads to the debye - waller - like suppression of the pancake oscillation amplitude and reduction of the corresponding current enhancement .
the pv displacements also modify the interaction force between pvs and jv lattice ( in the low orders with respect to displacements this effect is described by eq .
( [ iexpan ] ) ) . both above effects
have the same order and become strong when the typical pv fluctuation becomes of the order of the jv lattice period .
the pancake fluctuations also perturb phase distribution of the jv lattice leading to renormalization of its bare flux - flow resistivity .
we will also estimate this effect .
we now analyze the fluctuation effects quantitatively . at finite temperature and for the homogeneously moving jv lattice , @xmath49 ,
the pv displacements obey the following equation @xmath99 } { \partial \mathbf{u}_{n } } \!=\!\mathbf{f } _ { l , n}(t)\nonumber\end{aligned}\ ] ] where the function @xmath100 is defined by eqs .
( [ cdef ] ) and ( [ idef ] ) and @xmath101 is the thermal langevin force with the correlation function @xmath102 this equation generalizes eq .
( [ pancdynam ] ) for finite temperatures and large pancake displacements . introducing the dimensionless variables for the time , @xmath103 and coordinate @xmath104
, we conclude that the overall behavior is determined by the two dimensionless parameters , @xmath105 giving the relative strength of magnetic and josephson couplings between pvs in different layers and the reduced temperature @xmath106 this reduced temperature has a transparent physical meaning , at @xmath107 the fluctuating pancake displacement becomes of the order of the jv lattice period .
we can see that the relative strength of fluctuations rapidly increases with increasing in - plane field .
we numerically solved the dynamic langevin equations for the pancake displacements @xmath108 ( [ dyneq ] ) with numerically computed dependence @xmath38 from eq .
( [ idef ] ) , and used obtained solution to compute the average friction force ( [ frictforce ] ) .
figure [ fig : f_we_num]a shows the josephson - frequency dependences of the friction force @xmath109 for different reduced temperatures for typical value of the parameter @xmath110 , @xmath111 .
we can see that the main effect of fluctuations is suppression of the maximum of the @xmath112 dependence corresponding to the maximum of @xmath82 . to quantify this suppression ,
we show in fig .
[ fig : f_we_num]b temperature dependences of the maximum friction force for two values of the parameter @xmath113 , 0.1 and 0.2 .
the maximum force decreases by factor of 2 at @xmath114 .
the maximum only slightly shifts to higher frequencies with increasing temperature . also , at high temperatures the force decays slower than @xmath115 at large frequencies . ) for @xmath111 . in this plot @xmath109
is the friction force induced by one fluctuating pv stack on jv lattice defined by eq .
( [ frictforce ] ) in units of @xmath116 .
the frequency unit is @xmath117 .
the curves are calculated using pancake displacements obtained by direct numerical simulation of the langevin dynamic equations ( [ dyneq ] ) .
the upper solid line represents the analytical result ( [ pancfricforce ] ) valid at @xmath118 and @xmath119 .
( b)the temperature dependence of the maximum friction force for two values of the parameter @xmath113 , 0.1 and 0.2.,width=326 ] pancake fluctuations also perturb regular phase distribution in the moving jv lattice leading to modification of the regular flux - flow conductivity .
this gives an additional channel of pv influence on the jv mobility . in this section
we estimate this effect quantitatively .
dynamic behavior of the jv lattice in layered superconductors is described by the coupled reduced equations for the interlayer phase differences @xmath12 and reduced fields @xmath120 , see , e.g. , ref .
[ dyneqphase ] @xmath121 in these equations the units of space and time are given by the josephson length , @xmath122 , and the inverse plasma frequency , @xmath123 , @xmath124 and @xmath125 are the dissipation parameters related to the quasiparticle conductivities , and @xmath126 . for the moving triangular lattice the phase differences
can be represented as @xmath127 where @xmath128 is the phase perturbation induced by pv displacements , @xmath129 , of the stack located at @xmath130 , @xmath131 is the correction due to the josephson coupling . solving eqs.([dyneqphase ] ) by fourier transform in the first order with respect to @xmath132 , we obtain @xmath133 _ { \mathbf{k},\omega } } { \lambda(\mathbf{k},q,\omega)}\label{oscphase}\\ \text{with~}\lambda(\mathbf{k},q,\omega ) & = \!\omega^{2}\!-\!i\nu_{c}\omega \!-\!\frac{k^{2}(1\!+\!i\nu_{ab}\omega)}{2(1\!-\!\cos q)\!+\!\left ( 1\!+\!i\nu_{ab}\omega\right ) /l^{2}},\nonumber\end{aligned}\ ] ] where @xmath134 _ { \mathbf{k},\omega}$ ] notates fourier transformation with respect to coordinate and time . to obtain i - v characteristic we have to evaluate the average reduced josephson current @xmath135 , which in the first order with respect to @xmath131 is given by @xmath136 note that the pv excess current derived in the previous sections
also can be obtained using this approach if we add to @xmath131 the dynamic phase perturbation due to the oscillating pvs .
using result ( [ oscphase ] ) , we derive @xmath137 where @xmath138 is the fourier transform of the phase correlation function @xmath139 \right\rangle \\ & \approx\exp\left [ \!-\!\left\langle \left ( \theta_{v , n}(\mathbf{r},t ) \!-\!\theta_{v , n^{\prime}}(\mathbf{r}^{\prime},0)\right ) ^{2}\right\rangle\ ! /2\right]\end{aligned}\ ] ] and @xmath140.\ ] ] in particular , without pv fluctuations @xmath141 and the average josephson current is simply given by @xmath142 giving at small @xmath143 the bare jv flux - flow resistivity ( see ref . ) .
let us evaluate the phase correlation function @xmath144 .
neglecting correlations between displacements in different stacks , we obtain @xmath145 \right ) \label{pancphasecorr}\\ & g(\mathbf{r } , \mathbf{r } ^{\prime})\!=\!\frac{1}{2}\left\langle \left [ \theta_{p}(\mathbf{r},\mathbf{u}_{n+1}(t),\mathbf{u}_{n}(t))\!-\!\theta_{p}(\mathbf{r } ^{\prime},\mathbf{u}_{1}(0),\mathbf{u}_{0}(0))\right ] ^{2}\right\rangle\nonumber\end{aligned}\ ] ] using expansion of the pv phase difference at @xmath146 with @xmath147 , @xmath148 _ { z}/r^{2}$ ] , we can see that the integral over @xmath149 logarithmically diverges at large @xmath150 . cutting off this divergency at @xmath151 , we obtain with logarithmical accuracy the following result @xmath152 . from eqs .
( [ joscurrgen ] ) and ( [ pancphcorrest ] ) we can conclude that the pv fluctuations _ reduce _ the regular contribution to the josephson current and the relative reduction can be estimated as @xmath153 .
this correction roughly corresponds to suppression of the josephson coupling in the area @xmath154 around the pv stack . at small velocities
this reduction is typically much smaller than the current _ increase _ due to pancake oscillations .
however , as the pv excess current decays with increasing velocity , the regular contribution may become dominating at high velocities .
the relative contribution of the regular term also increases with increasing temperature .
columnar defects produced by heavy - ion irradiation are known to be the most efficient pinners of pvs , see ref . .
moreover , they effectively suppress fluctuation and align the pv stacks @xcite .
experimentally , this alignment leads to significant increase of the josephson plasma resonance frequency @xcite .
therefore we can expect that the columnar defects will strongly suppress the pv - induced damping of jvs .
naively , one may think that effect of the columnar defects can be described by a simple enhancement of the effecting spring constant in the confining parabolic potential .
however , a more accurate analysis below shows that this is incorrect description .
a columnar defect produces a very strong local potential for pvs , which is poorly described by the parabolic approximation . due to a discrete nature of the pv stack ,
its statistical distribution around the columnar defect consists of two parts @xcite : a sharp peak centered at the defect corresponding to pv located inside the column and a very wide envelope function corresponding to pv located outside the column .
in such situation , we can expect that pvs confined inside the column do not contribute to jv damping while pvs outside the column give almost the same contribution as free pvs .
this means that the columnar defect simply reduce the excess current by the probability factor to find pv outside the columnar defect , @xmath155 , as @xmath156 . to estimate this probability , we consider a fluctuating pancake vortex near the columnar defect , which we model as an insulating disk with radius @xmath157 .
the ratio of probability to find the pv outside the column , @xmath155 , to probability to find pv inside the column , @xmath158 , can be evaluated as @xmath159 where @xmath160 is the pv energy , measured with respect to the ground state corresponding to the pv located in the column center , @xmath161 with @xmath162 . the first term , @xmath163 is the pinning energy for @xmath164 , the second term is the interaction energy with columnar defect , and the third term is the magnetic coupling energy , @xmath165 is elemental area which can be evaluated by analyzing the gaussian fluctuation of the order parameter in the vicinity of vortex core @xcite , @xmath166 in which @xmath167 is the coherence length , @xmath168 is the ginzburg - landau parameter , and @xmath169 .
in contrast to consideration of ref .
, in the case of large in - plane field we can neglect josephson interaction in between pvs in neighboring layers and take into account only magnetic coupling .
the value of the integral ( [ relprobcd ] ) is determined by the competition between the two factors : the large energy cost @xmath170 for putting pv outside the column and large area of integration limited by the shallow magnetic - coupling parabolic potential . for estimate
, we can neglect the long - distance tail @xmath171 in the pinning potential . in this case
we obtain the following estimate @xmath172 therefore , the probability to find pv outside the column can be estimated as @xmath173 c_{s}/\kappa^{2}}\label{probresult}\ ] ] ratio @xmath174 becomes of the order of @xmath175 at the typical depinning temperature , @xmath176 , which can be estimated as @xmath177 the columnar defects strongly reduce the damping due to pvs at @xmath178 and this effect rapidly decreases at @xmath179 .
the value of this important temperature scale is most sensitive to the london penetration depth . for optimally
doped bscco , assuming @xmath180nm@xmath181 , @xmath182nm@xmath181 with @xmath183k , and @xmath184 nm , we estimate @xmath185k . for weakly underdoped
bscco , taking @xmath186 nm and @xmath187k , we obtain @xmath188k .
josephson flux - flow ( jff ) measurements have been done on the two mesa structures fabricated from the bscco single - crystal whiskers using double - sided processing by the focused ion beam ( fib ) technique @xcite , see fig .
[ fig : mesas ] .
the mesas had geometrical sizes l@xmath189l@xmath190l@xmath191=26@[email protected]@xmath192m@xmath194(0.18 - 0.23)@xmath192 m ( mesa # 1 ) and 20@xmath192m@xmath1945@xmath195m@xmath194(0.5 - 0.6)@xmath192 m ( mesa # 2 ) .
the mesa # 1 was prepared out of the overdoped whisker grown in oxygen flow .
the mesa # 2 was prepared out of the underdoped whisker grown in a mixture of 10 - 20 of o@xmath0 and 80 - 90% of ar . at fixed temperatures and fixed in - plane component of the magnetic field ( @xmath196 ) we measured the jv flux - flow resistance @xmath197 at several fixed currents and the i - v characteristics as a function of the c - axis component of the field ( @xmath198 c ) .
the cryostats with two perpendicular coils have been used for these experiments , the main coil provided the parallel field component , @xmath199 , while another coil induced the perpendicular component , @xmath200 . the strictly parallel orientation has been found using sharp maximum of @xmath201 dependence by sweeping @xmath200 .
we estimated an accuracy of field rotation in this way to be within 0.01@xmath202 .
the experiments have been carried out at high temperatures above 40k to avoid flux trapping effects . as a separate part of the experiments
, we studied the influence of columnar defects on the jff with and without the perpendicular field component .
the irradiation was made by pb - ions with energy of 1 gev at the ganil accelerator ( caen , france ) .
the defects were introduced into the mesa # 2 along the _ c _ axis with concentration @xmath203@xmath204 corresponding to the matching field of about 60 oe . for a comparison , the jff characteristics have been measured in the mesa # 2 first just before irradiation and then remeasured exactly at the same conditions within a week after irradiation .
dependence of the z - axis voltage at fixed current @xmath2050.2ma and in - plane filed @xmath206 for the mesa # 1 .
the right plot shows the current - voltage ( i - v ) dependences at several fixed @xmath200 values marked in the left plot .
the horizontal dashed line marks the current value in the left plot .
one can see that the i - v dependence acquires a jump around @xmath207oe , which is a coming from the nonmonotonic contribution , @xmath208 , from the pancake vortices . as a consequence ,
the @xmath200 ( or angular ) dependence of voltage at fixed current in the left side also has a jump .
, width=326 ] vs voltage for @xmath209 and three tilt angles demonstrating that the excess pancake current is proportional to concentration of pvs.,width=307 ] t and @xmath210 t , and for very close values of c - axis field @xmath211 and @xmath212oe . for reference , the i - v dependences at @xmath213oe
are also shown for both @xmath214 .
the amplitude of the excess pv current for @xmath215 t is more than four times smaller than for @xmath216t.,width=307 ] t and tilt angle 0.3@xmath202 corresponding to @xmath217oe . the voltage corresponding to maximum excess current decreases with decreasing temperature.,width=307 ] tilted at angle @xmath218 and @xmath219 ( corresponding to @xmath220oe ) .
irradiation dose corresponds to the matching field , @xmath221 g .
irradiation does not influence motion of the jv lattice without pvs , at @xmath222 but strongly reduces pancake effect at @xmath223k and @xmath224k .
this reduction is much weaker at @xmath225k.,width=268 ] figure [ fig : angscanivs](left panel ) shows the dependence of the jff voltage at fixed current on the _ c_-axis magnetic field , @xmath200 for the mesa # 1 .
we can see that the jff voltage is highly sensitive to a small component of @xmath200 resulting in large negative jff magnetoresistance @xmath226 . at the fixed current ( 0.2 ma ) and fixed in - plane field ( 1 tesla ) the jff resistance first decreases linearly with @xmath200 and then drops down sharply to the low - resistance state . to understand the mechanism of this drop
, we show in the right panel of fig .
[ fig : angscanivs ] a series of i - v dependences traced at the different fixed @xmath200 values marked in the left plot .
we can see from this plot , that the voltage drop to the low - resistance state originates from a nonmonotonic character of the i - v dependence which appears at high enough @xmath200 .
the excess current , @xmath227 , in the i - v dependences is a direct consequence of an additional damping of the moving jv lattice induced by the pv stacks , as discussed in the theoretical sections above .
figure [ fig : djhzscaling ] shows plots of the ratios @xmath228 vs bias voltage @xmath229 for the mesa # 2 before irradiation at different tilt angles corresponding to different @xmath200 .
one can see nonmonotonic dependence @xmath230 , that first increases rapidly , reaches a maximum and then decreases slowly .
another remarkable feature is proportionality of @xmath231 to @xmath200 up to @xmath232 oe ( 0.5@xmath202 tilt ) , which is seen from an almost ideal collapse of the @xmath228 curves at different tilt angles .
figure [ fig : mesa1hz40_5oe ] shows the i - v dependences of the mesa # 1 for two values of the in - plane field , @xmath216 t and @xmath210 t , and for very close values of c - axis field , @xmath233oe .
we can see that for @xmath215 t the amplitude of the excess current is more than four times smaller than for @xmath216 t , which is consistent with the theoretical estimates .
figure [ fig : ivstempcompare ] illustrates the temperature dependence of the excess current for the mesa # 2 .
we can see that at lower temperatures maximum in the excess current increases and moves to the lower voltages .
we also note that for the overdoped mesa the excess pv current density is 4 - 5 times larger than for the underdoped one .
figure [ fig : dicomp03degtemp ] shows comparison of the pv - induced damping for the mesa # 2 before and after irradiation for three temperatures , 40k , 50k , and 60k .
it is interesting to note that _ the columnar defects induced by heavy - ion irradiation have negligible direct influence on dynamics of jv lattice _
( ivs at @xmath218 tilt angles are not noticeably influenced by the irradiation ) . on the other hand
, they strongly influence the pv - induced damping at low temperatures .
we can see that at 40k and 50k pinning of the pv stacks by the columnar defects almost completely eliminates the pv extra damping . at 60 k this damping
is restored most probably due to the thermal depinning of the pv stacks from the columns .
this behavior is consistent with location of the thermal depinning temperature @xmath176 evaluated in sec.[sec : coldef ] ( for underdoped bscco we estimated @xmath234k ) .
strong dynamic interaction of the jvs with the pv stacks leads to a very high sensitivity of the jv transport to a very small concentration of the pv stacks which we observed experimentally . at a given driving force ( or dc current )
that interaction slows down the jv lattice and thus reduces the jff voltage leading to the pronounced negative jff magnetoresistance with respect to the _
c_-axis component of the magnetic field . on the other hand , at a given voltage
the extra damping due to the pv stacks results in increase of the current . theoretical description of these effects is based on calculation of the oscillating zigzag deformations of the pv stacks induced by the moving jv lattice and the average friction force due to these deformations .
several features of the experimental pv excess current , such as its nonmonotonic dependence on voltage , its decrease with increasing temperature and the in - plane field , are in a good _ qualitative _ agreement with the theory .
a strong suppression of the pv damping by columnar defects also provides confirmation for suggested mechanism .
a detailed _ quantitative _ comparison between the experiment and theory requires an accurate knowledge of the microscopic parameters of superconductor , london penetration depth , @xmath235 , anisotropy @xmath3 , and pancake flux - flow resistivity @xmath236 , as well as geometrical parameters of the mesas . to estimate the key theoretical parameters @xmath77 and @xmath78 we need values of the magnetic spring constant @xmath237 and the pv viscosity coefficient , @xmath58 .
the magnetic spring constant is mainly determined by the london penetration depth .
for example , taking a value @xmath238 nm typical for overdoped bscco at @xmath239k , we obtain @xmath240n / m@xmath241 and @xmath242 g .
the value of the viscosity coefficient is more uncertain . to our knowledge
no direct transport measurement of the pv flux - flow resistivity in bscco has been published .
microwave measurements of the flux - flow resistivity in single crystals ( at 40.8 ghz ) @xcite and films ( at 48 ghz ) @xcite show that @xmath243 has strong temperature dependence and suggest a typical value for @xmath244k , @xmath245@xmath246@xmath247-@xmath248n s / m@xmath241 corresponding to the slope @xmath249-@xmath250 @xmath251 cm / tesla .
this gives the following estimate for the flux flow - conductivity at @xmath252 , @xmath253@xmath7@xmath254-@xmath255[@xmath256cm]@xmath257 .
these data suggest that the maximum pv effect is expected at very small voltage , @xmath258-@xmath259v .
experimentally , the maximum excess current is observed at significantly higher voltages , @xmath260-@xmath261v corresponding to relaxation frequencies 7 - 15 ghz . also , the simple theory does not quite describe shapes of the experimental @xmath262 dependences , at @xmath263 the excess current decreases slower than expected @xmath264 decay . on the other hand
, we found that shapes of @xmath230 dependences also somewhat differ for underdoped and overdoped mesas .
these discrepancies can be explained if we ( i)assume small - velocity value for the viscosity constant , @xmath265n s / m@xmath266 , which is significantly larger than the microwave measurements suggest , and ( ii)assume that it decreases with increasing pv velocity . in the temperature and in - plane field region probed in experiment influence of thermal fluctuation pv s considered in sec .
[ sec : pvfluct ] is significant .
for example , for the overdoped mesa at 65 k , we estimate @xmath267-@xmath268 and the reduced temperature given by eq .
( [ redt ] ) as @xmath269 for @xmath270 t and @xmath271 for @xmath272 t .
shape deviations , at least partly , can be due to the pv thermal fluctuations .
our calculations show that the pv fluctuations mainly reduce the amplitude of @xmath231 and somewhat weaken the dependence @xmath230 at large @xmath229 but they do not shift much the maximum of the @xmath230 dependence .
another possible reason for disagreement with the simple theoretical picture is that at small velocities the jv lattice may move inhomogeneously , via thermally - activated jumps .
this regime of motion is not considered by the theory at all and it may lead to an increase of apparent @xmath273 . we expect that this inhomogeneous regime is more pronounced for smaller sample sizes in which averaging over pv locations is incomplete .
therefore , a systematic size dependence of the effect would help to understand mechanism of the low - voltage regime .
note that , due to a very small value of @xmath78 , without scrutinizing i - v shape , the low - voltage enhancement of damping can be easily misinterpreted as an enhancement of the critical current . in our experiments ,
a detailed study of the small - voltage region shows that the jv lattice slowly moves below the maximum current meaning that this maximum is due to the enhancement of damping .
this , of course , does not exclude possibility that in other experimental conditions ( e.g. , at low temperatures ) the main pv effect would be enhancement of the critical current .
an extreme sensitivity of the jv flux - flow voltage to a very small concentration of the pvs suggests a very attractive possibility to use small - size mesas for detection of penetration of individual pv stacks which will be seen as current or voltage jumps .
the change of the total current due to the penetration event of a single pv stack , @xmath274 , is expected to be independent on the mesa size and is given by @xmath275 its maximum value , @xmath276 , is realized at @xmath277 and at low temperatures , where thermal - fluctuation suppression is small , can be estimated as @xmath278 taking @xmath279a / cm@xmath241 , @xmath281 nm , and @xmath282 , we estimate @xmath283a . on the other hand , for the underdoped mesa # 2 at 50k from fig .
[ fig : djhzscaling ] we obtain @xmath284a consistent with the above theoretical estimate . for the overdoped mesa @xmath276
is 4 - 5 times larger .
such currents are detectable experimentally .
we conclude that this technique can be practically used for detection of the individual pv stacks penetration .
we would like to thank a. m. nikitina for providing the bscco single - crystal whiskers , a. ryhd , g. karapetrov , and m. gaifullin for valuable technical assistance .
aek also would like to thank k.kadowaki and y. kakeya for useful discussions .
work in argonne was supported by the u.s .
doe , office of science , under contract # w-31 - 109-eng-38 .
aek and yil acknowledge support from the nato travel grant no .
the work in ecole polytechique was partially performed in the frame of the cnrs - ras associated european laboratory between crtbt and ire `` physical properties of coherent electronic states in condensed matter '' .
efetov , zh .
. fiz . * 76 * , 1781 ( 1979 ) [ sov .
jetp * 49 * , 905 ( 1979 ) ] ; a. i. buzdin and d.feinberg , _ j. phys .
( paris ) _ * 51 * , 1971 ( 1990 ) ; s. n. artemenko and a. n. kruglov , phys .
a * 143 * 485 ( 1990 ) ; j. r. clem , phys .
b * 43 * , 7837 ( 1991 ) .
l. n. bulaevskii and j. r. clem , phys .
b , * 44 * , 10234 ( 1991 ) .
l. n. bulaevskii , m. ledvij , and v. g. kogan , phys .
b * 46 * , 366 ( 1992 ) .
a. e. koshelev , phys
. rev .
lett . * 83 * , 187 ( 1999 ) . c. a. bolle ,
p. l. gammel , d. g. grier , c. a. murray , d. j. bishop , d. b. mitzi and a. kapitulnik , phys
. rev .
* 66 * , 112 ( 1991 ) .
t. matsuda , o. kamimura , h. kasai , k. harada , t. yoshida , t. akashi , a. tonomura , y. nakayama , j. shimoyama , k. kishio , t. hanaguri , and k. kitazawa , science , * 294 * , 2136(2001 ) .
i. v. grigorieva , j. w. steeds , g. balakrishnan , and d. m. paul , phys .
b * 51 * , 3765 ( 1995 ) .
a. grigorenko , s. bending , t. tamegai , s. ooi , and m. henini , nature * 414 * , 728 ( 2001 ) . v. k. vlasko - vlasov , a. e. koshelev , u. welp , g. w. crabtree , and k. kadowaki , phys . rev .
b * 66 * , 014523 ( 2002 ) .
m. tokunaga , m. kobayashi , y. tokunaga , and t. tamegai , phys .
b * 66 * , 060507(r ) ( 2002 ) .
s. bending and m. j. w. dodgson , j. phys .
: condens .
matter * 17 * , r955 ( 2005 ) .
j. u. lee , j. e. nordman , and g. hohenwarter , appl .
phys . lett . ,
* 67 * , 1471 ( 1995 ) ; j. u. lee , p. guptasarma , d. hornbaker , a. el - kortas , d. hinks , and k. e. gray appl .
lett . , * 71 * , 1412 ( 1997 ) .
g. hechtfischer , r. kleiner , a. v. ustinov , and p. mller , phys .
lett . * 79 * , 1365 ( 1997 ) ; g. hechtfischer , r. kleiner , k. schlenga , w. walkenhorst , p. mller , and h. l. johnson , phys . rev .
b , * 55 * , 14638 ( 1997 ) .
i. latyshev , p. monceau , and v. n. pavlenko , physica c * 282 - 287 * , 387 ( 1997 ) ; physica c * 293 * , 174 ( 1997 ) ; yu .
i. latyshev , m. b. gaifullin , t. yamashita , m. machida , and y. matsuda , phys .
lett . , * 87 * , 247007 ( 2001 ) .
l. n. bulaevskii , m. maley , h. safar , and d. domnguez , phys .
b * 53 * , 6634 ( 1996 ) .
l. civale , a. d. marwick , t. k. worthington , m. a. kirk , j. r. thompson , l. krusin - elbaum , y. sun , j. r. clem , and f. holtzberg , phys .
lett . , * 67*(5 ) , 648 ( 1991 ) ; m. konczykowski , f. rullier - albenque , e. r. yacoby , a. shaulov , y. yeshurun , and p. lejay , phys . rev .
b , * 44 * , 7167 ( 1991 ) ; w. gerhuser , g. ries , h. w. neumller , w. schmidt , o. eibl , g. saemann - ischenko , and s. klaumnzer , phys . rev .
lett . , * 68 * , 879 ( 1992 ) ; l. civale , supercond .
* 10 * a11 ( 1997 ) ; c. j. van der beek , m. konczykowski , r. j. drost , p. h. kes , n. chikumoto , and s. bouffard , phys .
b * 61 * , 4259 ( 2000 ) .
m. sato , t. shibauchi , s. ooi , t. tamegai , and m. konczykowski , phys .
lett . * 79 * , 3759 ( 1997 ) ; m. kosugi , y. matsuda , m. b. gaifullin , l. n. bulaevskii , n. chikumoto , m. konczykowski , j. shimoyama , k. kishio , k. hirata , and k. kumagai , phys .
lett . * 79 * , 3763 ( 1997 ) | we study theoretically and experimentally influence of pancake vortices on motion of the josephson vortex lattice in layered high - temperature superconductors .
mobility of the josephson vortices in layered superconductors is strongly suppressed by small amount of pancake - vortex stacks . moving josephson vortex lattice forces oscillating zigzag deformation of the pancake - vortex stacks contributing to damping .
the salient feature of this contribution is its nonmonotonic dependence on the lattice velocity and the corresponding voltage .
maximum pancake effect is realized when the josephson frequency matches the relaxation frequency of the stacks .
the pancake - vortex damping is strongly suppressed by thermal fluctuations of the pancake vortices .
this theoretical picture was qualitatively confirmed by experiments on two mesas prepared out of bi@xmath0sr@xmath1cacu@xmath1o@xmath2 whiskers .
we found that the josephson - vortex flux - flow voltage is very sensitive to small c - axis magnetic field .
the pancake - vortex contribution to the current indeed nonmonotonically depends on voltage and decreases with increasing temperature and in - plane magnetic field .
we also found that irradiation with heavy ions has no noticeable direct influence on motion of the josephson vortices but dramatically reduces the pancake - vortex contribution to the damping of the josephson vortex lattice at low temperatures . |
the point - like x - ray source 2xmmj011028.1@xmath6460421 ( henceforth , hlx-1 for simplicity ) is the strongest intermediate - mass black hole ( imbh ) candidate known to date @xcite .
it is seen in the sky at a distance of @xmath7 from the nucleus of the s0 galaxy eso243 - 49 ( redshift @xmath8 , luminosity distance @xmath9 mpc , distance modulus @xmath10 mag ; at this distance , @xmath11 pc ) .
its x - ray luminosity and spectral variability @xcite and its radio flares detected in association with the x - ray outbursts @xcite are consistent with the canonical state transitions and jet properties of an accreting bh . with a peak x - ray luminosity @xmath12 erg s@xmath13 , the bh mass required to be consistent with the eddington limit is @xmath14 .
a similar value is obtained from spectral modelling of the thermal x - ray component , which is consistent with emission from an accretion disc @xcite . if these bh mass estimates are correct , hlx-1 is way too massive to have been formed from any stellar evolution process .
a more likely scenario is that it is the nuclear bh ( perhaps still surrounded by its own nuclear star cluster ) of a disrupted dwarf satellite galaxy , accreted by eso243 - 49 @xcite .
hlx-1 has a point - like , blue optical counterpart ( @xmath15 mag near the outburst peak ; @xcite ) .
the presence of h@xmath16 emission at a redshift consistent with that of eso243 - 49 @xcite is perhaps the strongest argument for a true physical association .
it is still debated whether the optical continuum emission is dominated by the outer regions of the bh accretion disc , or by a young star cluster around the bh @xcite . in the absence of phase - resolved dynamical measurements of the bh motion
, we can use the _ swift _ x - ray lightcurve properties to constrain the system parameters .
the x - ray flux shows recurrent outbursts every @xmath17 d ( seen every late august in 2009 , 2010 , 2011 and 2012 ) , due either to some kind of disc instability , or to a periodic enhancement of the accretion rate .
several alternative scenarios were considered and discussed by @xcite , who favoured a model in which enhanced mass transfer into a quasi - permanent accretion disc is triggered by the passage at periastron of an asymptotic giant branch ( agb ) star on an eccentric orbit ( @xmath18 ) . since the publication of that work , the detection of the third and fourth consecutive outbursts ( see @xcite for the first report of this year s outburst ) has clinched the interpretation of the recurrence timescale as the binary period . furthermore , additional optical photometric results have been published , based on data from the _ hubble space telescope ( hst ) _
@xcite and from the european southern observatory ( eso ) s very large telescope ( vlt ) @xcite .
thus , in this paper we revisit and update @xcite s orbital models and constraints in the light of the new results .
at a distance of 95 mpc , the characteristic size of the region responsible for most of the soft , thermal ( @xmath19 kev ) x - ray emission is @xmath20 a few @xmath21 cm ( inferred from fits to _ xmm - newton _ , _ chandra _ , _ swift _ spectra ) , and is consistent with being constant during the decline of individual outbursts , and over the three recorded outbursts @xcite .
this suggests that the soft x - ray emission traces the true inner radius of the disc , bounded by the innermost stable circular orbit around the bh . instead ,
much less is known about the outer disc radius , from uv / optical / ir observations ; it is still debated how much of the blue optical emission comes from an irradiated disc , and how much from a possible cluster of young stars around the bh . if the disc is the dominant uv / optical emitter , the _ hst _ and vlt studies of @xcite and @xcite , respectively , agree on an outer disc radius @xmath22 cm @xmath23 au cm , for a viewing angle @xmath24 ; the radius is smaller for a face - on view , which we consider a more likely scenario , given the narrow full - width - half - maximum of the h@xmath16 emission line ( work by hau & soria , in preparation ) .
even if we adopt this slightly larger upper limit for the disc radius , the argument for a high eccentricity discussed in section 3 remains unchanged . ] .
if a substantial contribution comes from unresolved young stars , we can take that value as an upper limit to the true disc size .
a ratio of outer / inner disc radii @xmath25 is significantly smaller than observed in transient galactic bhs with roche - lobe - filling donors , where typical outer radii are @xmath26 cm @xmath20 a few @xmath27 times the innermost stable circular orbit @xcite .
this serves as a warning that we have to disentangle what scales with bh mass and what does not , when using scaled - up galactic bh models to interpret hlx-1 .
while the inner disc depends directly on the bh mass , the outer disc depends mostly on the donor star and binary separation .
there is an alternative way to estimate the outer disc size , based on the x - ray outburst decline timescale .
following @xcite and @xcite , we assume that the outbursting disc is approximately in a steady state with surface density @xmath28 where @xmath29 is the central accretion rate and @xmath30 the kinematic viscosity .
when the whole disc from @xmath31 to @xmath32 is in a hot , high - viscosity state , the total mass in the disc @xmath33 where we have neglected other sources of mass loss from the disc apart from bh accretion . in eq.(2 ) , @xmath30 is interpreted as an average value of the kinematic viscosity over the whole disc ; in practice , we take the value of @xmath30 near the outer edge of the disc @xcite . integrating eq.(2 ) , we obtain the well - known exponential decline for the disc mass @xmath34 and consequently also for the accretion rate @xmath35 and the outburst luminosity @xmath36 . in summary , we expect to see a luminosity @xmath37 where @xmath38 is the value at the outburst peak , declining on a timescale @xmath39 as long as the rate at which the disc mass is depleted during the outburst decline is much larger than any ongoing transfer of mass from the donor star . for the viscosity at the outer edge of the disc
, we take the usual parameterization @xmath40 @xcite , where @xmath16 is the viscosity coefficient in the hot state , @xmath41 is the sound speed and @xmath42 the vertical scaleheight . in the simplest , order - of - magnitude approximation
, we can take an outer disc temperature @xmath43 k ( enough to keep it in the hot state ) , corresponding to @xmath44 cm s@xmath13 , and a vertical height @xmath45 .
this gives , from eq.(6 ) : @xmath46 which we shall directly compare with the observations .
if we adopt the shakura - sunyaev disc solution @xcite , with kramers opacity , we can write @xmath47 where @xmath48 is the accretion rate in units of @xmath49 g s@xmath13 , @xmath50 is the bh mass in solar units , @xmath51 , etc . then , from eq.(6 ) : @xmath52 that is @xmath53 the peak luminosity @xmath38 in eq.(2 ) can be left as a purely observational parameter , or it can itself be expressed as a function of outer disc radius , viscosity , density and bh mass , if we assume that the outburst is triggered via the dwarf - nova instability @xcite .
( more precisely , if we assume that the outburst starts when the enhanced mass transfer due to periastron passage pushes the disk from the cold to the hot state . ) in that case , the surface density at any radius immediately before the start of the outburst approaches the maximum value allowed by the s - curve in the surface - temperature phase space @xcite : @xmath54 where @xmath55 is the viscosity parameter in the cold disc state . taking for simplicity @xmath56 , where @xmath57 is a constant @xmath58
, we can then express the maximum volume density at the start of the outburst as @xmath59 , which is essentially independent of @xmath60 , given the expression for @xmath61 in eq.(11 ) .
it is then easy to integrate the total disc mass at the start of the outburst : @xmath62 the disc mass at later times @xmath63 the accretion rate ( cf .
eq.(4 ) ) @xmath64 and the peak luminosity @xmath65 where @xmath61 comes from eq.(11 ) .
so far , we have assumed that the whole disc is in the hot state ; this is usually the case in the early part of an outburst , especially when the outer edge of the disc is kept in the hot state by x - ray irradiation .
the exponential decay continues until the outer disc annuli can no longer be kept in the hot state , so that hydrogen recombines and viscosity drops . from that moment , the outer edge of the hot disc @xmath66 .
the contribution to the accretion rate and to the continuum x - ray / uv / optical emission from the outer ( cold , low - viscosity ) annuli at @xmath67 becomes negligible .
it was shown by @xcite that the central accretion rate in this second phase of the decline is @xmath68,\ ] ] where @xmath69 is the time after which the outer disc is no longer in the hot state , and @xmath70 parameterizes the fraction of x - ray luminosity intercepted and thermalized in the outer disc . as @xmath70 can be taken as a constant , eq.(16 ) shows that the accretion rate and luminosity decline in the late part of the outburst is linear .
most importantly for our current purpose , the slope of the linear decline is such that @xmath71 @xcite , where @xmath72 is the ( extrapolated ) time in which the accretion rate and luminosity go to zero . finally , we need to consider the case when there is ongoing mass transfer @xmath73 from the donor star during the outburst . in that case ,
the asymptotic value of the luminosity in the exponential decline is not zero but @xmath74 ( assuming a standard radiative efficiency @xmath75 ) .
recalling that @xmath69 is the time when the lightcurve switches from an exponential to a linear decline , and defining @xmath76 , eq.(5 ) is modified as @xcite @xmath77 after the transition to a linear regime , the luminosity is @xmath78.\ ] ] note that if @xmath79 , the first derivative of the luminosity is discontinuous at @xmath80 @xcite , because the gradient of the exponential decay is @xmath81 that is flatter than the gradient of the linear decay @xmath82 therefore , the exponential - to - linear transition is often referred to as the `` knee '' in the lightcurve of transient x - ray binaries .
we shall now fit the x - ray lightcurve to obtain two independent estimates of the viscous timescale @xmath83 , from the exponential and the linear regime , and use them to constrain @xmath32 from eq.(10 ) , or , using a simpler approximation for the scale - height , from eq.(7 ) .
we shall then derive an independent estimate of @xmath32 from the expression for the peak luminosity in eq.(15 ) .
we studied the publicly available _ swift _ x - ray telescope ( xrt ; * ? ? ? * ) data for the 2010 outburst , because it is the same outburst for which we obtained constraints on @xmath32 from the optical continuum @xcite .
we used the on - line _ swift_/xrt data product generator @xcite to extract a lightcurve in the @xmath84@xmath85 kev band .
we fitted the lightcurve with an initial exponential decay , a knee and a linear decay ( fig . 1 ) .
the shape of the x - ray outburst lightcurve of hlx-1 is remarkably similar to those of several transient galactic x - ray binaries ( bhs and neutron stars ) , modelled by @xcite , which were successfully used to constrain the size of their accretion discs . for the exponential part
, we obtain a best - fitting timescale @xmath86 s ( @xmath87 confidence limit ) .
for the linear part , we have @xmath88 s. assuming a peak luminosity @xmath12 erg s@xmath13 in the @xmath84@xmath85 kev band , and a bolometric luminosity a factor of 2 higher , implies an accretion rate @xmath89 g s@xmath13 at standard efficiency . the viscosity parameter @xmath90 , and more likely @xmath91 @xcite . from eq.(10 ) , this implies @xmath92 cm , only very weakly dependent on bh mass and accretion rate . using the approximation in eq.(7 ) , we also obtain @xmath93 cm . from the peak luminosity ( eq.(15 ) ) , for @xmath91 we obtain @xmath94 cm .
we also analyzed the lightcurves for the 2009 and 2011 outbursts ( the 2012 outburst was still ongoing as this paper went to press ) .
they are more noisy , less easy to interpret in terms of exponential and linear branches . however , for both of them it is possible to estimate an e - folding decline timescale , roughly corresponding to the exponential timescale determined for the 2010 outburst .
the timescales are @xmath95 s and @xmath96 s for 2009 and 2011 , respectively , and the peak luminosities are approximately the same in all three outbursts .
thus , we also estimate an outer radius @xmath97 cm in the 2009 and 2011 outbursts , in the standard disc approximation .
those values are almost one order of magnitude smaller than what we estimated by assuming that most of the continuum emission comes from the hot disc ; and the latter was already a surprisingly small radius compared with the binary system parameters ( see section 3 ) .
we note that both the _ hst _ and vlt observations @xcite were taken during the exponential part of the decline , that is when the whole disc was in a hot state .
therefore , the estimates of @xmath32 from the optical / uv continuum should be comparable to those from the x - ray lightcurve .
new optical observations early in the next outburst will hopefully allow us to measure the true outer disc size and luminosity .
we can now compare the size of the disc estimated from optical and x - ray flux measurements ( @xmath0 cm @xmath1 cm ) with the characteristic size of the binary system . because of the sharpness of the outbursts rise and decline , reminiscent of galactic x - ray binaries , we assume that the bh is accreting from a single donor star rather than agn - like gas inflows .
we also assume that the outburst recurrence timescale @xmath98 d corresponds to the binary period .
then , the semimajor axis @xmath99 of the binary is @xmath100 @xcite , where @xmath101 and @xmath102 .
typical values for hlx-1 in the intermediate - mass bh scenario are @xmath103 and @xmath104@xmath105 . therefore , in the most accepted scenario , the semimajor axis is at least 10 , and possibly up to 100 times larger than the disc radius .
this mismatch clearly suggests an eccentric orbit @xcite , in which the characteristic disc size is determined by the periastron separation @xmath106 , with eccentricity @xmath107 .
the amount of mass transferred to the bh in each outburst suggests that the donor star overflows its instantaneous roche lobe every time it passes at periastron . in roche - lobe mass transfer systems , the outer edge of the accretion disc
is generally identified with the largest stable non - intersecting orbit ( tidal truncation radius @xmath108 ) . for mass ratios
@xmath109 , @xmath110 @xcite .
if unstable orbits are also allowed , the disc may expand up to @xmath111 ( * ? ? ? * and references therein ) .
if we take the periastron distance as the instantaneous binary separation during the phase of roche - lobe overflow , this corresponds to an expected disc size @xmath112 . for an observed disc size @xmath113 cm
, the tidal radius condition would require an eccentricity @xmath114 for a bh mass @xmath115 , and @xmath116 for a bh mass @xmath117 .
if we take the lower bound to our observed disc size @xmath118 cm , we need @xmath119 or @xmath120 , respectively .
we can argue that the tidal truncation constraint is not relevant to the case of hlx-1 , where mass transfer may occur impulsively near periastron , and the timescale for the disc to expand to its tidal truncation radius is similar to the timescale for the disc matter to be accreted and for the binary orbit to expand after periastron . in other words , in hlx-1 the disc may look small because it did not have time to grow to its tidal truncation radius .
instead , the circularization radius provides a stronger lower limit to the predicted disc size , and is applicable to any system where mass transfer occurs through the lagrangian point l@xmath121 .
the circularization radius @xmath122 is defined via the conservation of angular momentum equation @xmath123 where @xmath124 is the orbital velocity of the accretion stream around the bh , @xmath125 the distance between the bh and the lagrangian point l@xmath121 , and @xmath126 the angular velocity of the donor star at periastron . here
, we must be careful not to use the well - known fitting formula for @xmath127 @xcite because it applies only for @xmath128 and for circular orbits .
instead , we need to compute @xmath129 from eq.(a13 ) of @xcite , valid for eccentric orbits : @xmath130 here , @xmath131 is the ratio of the rotational angular velocity of the donor star to its orbital angular velocity , at periastron .
it parameterizes the degree of tidal locking ; a star with @xmath132 is rotating synchronously with the orbit , at periastron . a list of @xmath129 solutions for characteristic values of @xmath133 and @xmath134 is given in table 1 .
the orbital velocity of the accretion stream around the bh @xmath135 and the orbital angular velocity at periastron @xmath136^{1/2}.\ ] ] for the sake of our numerical estimate , we take @xmath137 , a typical value in the range of parameters thought to be relevant for hlx-1 ( table 1 ) . then , substituting into eq.(23 ) , we have : @xmath138 assuming that the bh mass is in the range @xmath25@xmath139 , eq.(27 ) gives us a strong constraint on @xmath134 , by imposing that the circularization radius is smaller than the observed outer disc size . for example , assuming @xmath140 , a circularization radius @xmath141 cm ( at the upper end of our disc size estimates ) requires @xmath142 ; for @xmath143 cm ( at the lower end of our disc size estimates ) , @xmath144 .
the corresponding periastron distances , the periastrion distance between the two stars is always smaller than the circularization radius around the accreting primary .
this is because the angular momentum of the secondary at periastron , and of the matter transiting through the l@xmath121 point , is larger than the angular momentum of a circular orbit with the same binary separation .
it is not a problem , because by the time the accretion stream has completed a full orbit around the bh and formed a ring , the secondary has moved away from periastron and the primary s roche lobe has widened . ]
@xmath145 cm for @xmath146 , and @xmath147 cm for @xmath148 .
lcccc + & @xmath149 & @xmath150 & @xmath151 & @xmath152 + + @xmath153 & + @xmath154 & + @xmath155 & + @xmath156 & + + + @xmath153 & @xmath157 & @xmath158 & @xmath159 & @xmath159 + @xmath154 & @xmath160 & @xmath161 & @xmath161 & @xmath161 + @xmath155 & @xmath162 & @xmath163 & @xmath163 & @xmath163 + @xmath155 & @xmath164 & @xmath165 & @xmath165 & @xmath165 + the distance between l@xmath121 and the centre of the donor star places an upper limit on its radius ; characteristic values of its instantaneous volume - averaged roche lobe radius at periastron can be obtained from @xcite . for typical values @xmath166@xmath167 ,
the secondary gets squeezed to a radius @xmath168@xmath169 .
since we have assumed that the secondary fills the roche lobe and dumps mass into the bh only near periastron , this radius must also be similar to the size of the donor star .
for example , if @xmath170 and @xmath171 , the donor star must have a radius @xmath172 , consistent with main sequence and subgiant stars .
the values of @xmath134 estimated here from disc size arguments are much more extreme than what was suggested in @xcite .
they may seem implausible , knowing that tidal forces tend to circularize orbits in x - ray binaries .
however , @xcite showed that in the case of a donor star that transfers mass impulsively only at periastron , with @xmath173 , the secular evolution of the orbit leads to an increase of both eccentricity and semimajor axis , even when the opposite effect of tidal forces is taken into account .
if tidal circularization is neglected , the eccentricity increases as @xmath174 and the semimajor axis ( and hence the binary period ) as @xmath175 it is then easy to show from eq.(28,29 ) that the periastron distance @xmath176 , and therefore also the size of the secondary s roche lobe at periastron , remains unchanged . finally , we need to assess what donor stars can survive on such eccentric orbits with small periastron distance , avoiding tidal disruption .
the condition for survival is that the periastron distance @xmath177 is larger than the tidal disruption radius @xcite @xmath178 by substituting @xmath179 into eq.(31 ) , we can re - cast the tidal survival condition as @xmath180 easily satisfied in the likely mass range of hlx-1 .
we compared the estimates of the disc size from the optical continuum flux ( @xmath181 cm ) with those obtained by fitting the x - ray luminosity decline after an outburst .
we found that , at least for the 2010 outburst , the decline displays the standard sequence of exponential phase , knee , linear phase often seen in galactic x - ray binaries ; this strengthens the interpretation that the thermal x - ray emission in hlx-1 comes from a disc , and the decline timescale corresponds to its viscous timescale .
for all three outbursts observed to - date , the timescale is very short , @xmath182@xmath183 weeks .
this is similar or only slightly longer than what is typically observed in galactic x - ray transients , despite that fact that both the orbital period and the bh mass ( and , hence , the semimajor axis ) of hlx-1 are @xmath184 times larger .
the outer disc radius estimated from the viscous timescale is @xmath93 cm , if the viscosity parameter is similar to the values usually estimated for galactic bh transients in a high state .
we can not rule out that the fast accretion of the disc matter in hlx-1 may be partly due to an effective viscosity @xmath185 , higher than in the shakura - sunyaev prescription .
but we argue that even if we assume the upper disc size estimate @xmath186 cm , a highly eccentric orbit is required to explain the small disc size . to quantify the eccentricity , we calculated the characteristic length - scales of the binary system , as a function of bh mass and eccentricity . if the disc extends at least as far as the circularization radius ( as is usually the case in x - ray binaries with roche - lobe mass transfer )
, we obtain that @xmath187 , and therefore @xmath4 for a bh mass @xmath188 .
we argued that x - ray binaries with such extreme values of @xmath134 are the most likely evolutionary endpoint of systems with @xmath189 and a moderately eccentric initial orbit , such that roche - lobe - overflow mass transfer occurs only impulsively near periastron .
secular evolution will tend to make the orbit more and more eccentric , by increasing the semimajor axis and the binary period , at constant periastron distance .
the small periastron distance required to explain the hlx-1 observations sets an upper limit to the current radius of the donor star @xmath190 a few @xmath191 , ruling out supergiants , red giants and agb stars .
possible donors are main sequence ( b type or later ) or subgiants .
the compactness of the donor star , and the fact that secular orbital evolution due to mass transfer will not change the periastron distance , imply that the companion star in hlx-1 is not at immediate risk of tidal disruption , and will not be in the near future .
in other words , we are not observing hlx-1 in a peculiar moment of its evolution , immediately prior to tidal break - up of the donor star .
hlx-1 appears to be a stable system , with a lifetime for x - ray outbursts determined primarily by the mass transfer timescale from the donor ; at a rate @xmath192 yr@xmath13 ( averaged over the binary period ) , it may last for another @xmath193@xmath194 yr , during which its semimajor axis and binary period ( and hence , interval between outbursts ) will continue to increase .
eccentricities @xmath195 may seem implausibly extreme , but there is at least one class of stellar objects where they are the norm : s stars observed on highly eccentric orbits within 0.01 pc of the galactic nuclear bh @xcite .
a possible scenario for the origin of galactic s stars is the tidal disruption of a stellar binary system near the bh , which produces an escaping , hyper - velocity star , and a more tightly bound star on a very eccentric orbit , theoretically as high as @xmath196 @xcite .
observationally , the most eccentric , bound s star for which orbital parameters have been reliably determined has @xmath197 @xcite .
we speculate that intermediate - mass bhs in star clusters may also capture stellar companions on very eccentric orbits through a similar process .
i thank guillaume dubus , sean farrell , jeanette gladstone , pasi hakala , michela mapelli , george hau , albert kong , tom russell , luca zampieri for insightful discussions on the nature of hlx-1 .
this work was improved also thanks to the feedback received from several colleagues after my presentation at the iau symposium 290 in beijing . | we compare the outer radius of the accretion disc in the intermediate - mass black hole candidate hlx-1 as estimated from the uv / optical continuum , with the values estimated from its outburst decline timescales .
we fit the _ swift _ 2010 outburst decline lightcurve with an exponential decay , a knee and a linear decay .
we find that the disk has an outer radius @xmath0 cm @xmath1 cm , only an order of magnitude larger than typical accretion discs in the high / soft state of galactic black holes .
by contrast , the semimajor axis is @xmath2 a few @xmath3 cm .
this discrepancy can be explained with a highly eccentric orbit .
we estimate the tidal truncation radius and circularization radius around the black hole at periastron , and impose that they are similar or smaller than the outer disk radius .
we obtain that @xmath4 , that the radius of the donor star is @xmath5 a few solar radii , and that the donor star is not at risk of tidal disruption .
if the companion star fills its roche lobe and impulsively transfers mass only around periastron , secular evolution of the orbit is expected to increase eccentricity and semimajor axis even further .
we speculate that such extremely eccentric systems may have the same origin as the s stars in the galactic centre .
[ firstpage ] accretion , accretion discs x - rays : individual : hlx-1 black hole physics . |
on 2008 january 9 , a luminous x - ray transient was serendipitously discovered in ngc 2770 during the follow - up observation of sn 2007uy in the same galaxy with the _ swift _ satellite @xcite .
an optical counterpart was also discovered at the position of the transient @xcite , and it was named supernova ( sn ) 2008d @xcite . the total energy emitted at x - ray wavelengths is @xmath4 ergs , smaller than that of long gamma - ray bursts ( grbs ) by a factor of @xmath5 @xcite .
the origin of the x - ray emission is being debated .
@xcite , @xcite , and @xcite interpreted the x - ray transient as a sn shock breakout . on the other hand , @xcite , @xcite and @xcite suggested that the transient is the least energetic end of grbs or x - ray flashes .
sn 2008d is classified as type ib because of the presence of he lines @xcite while sne associated with grbs are all type ic ( without he lines ) .
the progenitor star of sn 2008d has a he layer prior to the explosion , and the he core mass is estimated to be 6 - 8 @xmath2 @xcite . in this paper
, we present optical spectroscopic and photometric observations of sn 2008d at @xmath6 300 days after the explosion ( nebular phases ) with the subaru telescope equipped with focas @xcite and the very large telescope ( vlt ) equipped with fors1 @xcite .
we find that the spectrum of sn 2008d shows a double - peaked [ ] emission profile , suggesting that sn 2008d is a bipolar explosion viewed from near the equatorial direction .
section of the @xmath7-band image centered on sn 2008d taken with subaru telescope at @xmath8 days .
the position of the sn is shown with the arrow .
two rectangles show the regions where the galaxy spectrum is extracted .
[ fig : image ] ] a spectroscopic observation of sn 2008d was performed on 2009 january 6 ( @xmath9 days ) with the subaru telescope .
hereafter @xmath10 denotes the epoch in observer s frame measured from the explosion date , mjd=54474.56 @xcite .
blue ( 4800 - 8000 ) and red ( 6000 - 9000 ) spectra were taken separately .
we used a slit of @xmath11 width and two 300 lines mm@xmath12 grisms .
the slit was placed with the position angle of @xmath13 ( north - south ) .
typical seeing during the observation was fwhm @xmath14 measured with star profiles in @xmath7 band .
no filter was used for the blue spectrum , while the o58 filter was used for the red spectrum to eliminate second order light .
the exposure time is 9000 and 3300 sec for the blue and red spectra , respectively .
the wavelength resolution is @xmath15 .
the observed data were bias - subtracted and flat - corrected , and then two - dimensional spectra were extracted .
wavelength calibration was performed with a th - ar lamp .
since the sn is located inside a brightness trough which surrounds the sn position with a diameter of @xmath16 arcsec ( see figure [ fig : image ] ) , simultaneous subtraction of the background galaxy and sky lines using the neighboring region leads to negative sn flux .
thus , we first extracted a one - dimensional galaxy spectrum using the region at 3 - 12 arcsec north and south from the sn position ( marked with boxes in figure [ fig : image ] ) .
then , the galaxy spectrum was subtracted using the spatial profile at 6920 - 6990 , in which the sn does not show any emission line . to maximize s /
n of the sn spectra especially at emission lines , we extracted the galaxy spectrum using a wide region .
since the global color of the galaxy spectrum depends on the position somewhat , this leads to a slight over - subtraction in the extracted sn spectrum in the blue ( see figure [ fig : spec ] ) .
however , the profile of the sn emission lines are not affected by the choice of integrated region for the galaxy spectrum .
the background - subtracted spectra were summed into one - dimensional spectra , and then , the flux was calibrated using the spectrophotometric standard star feige 34 @xcite .
the blue and red spectra were combined , weighted by the exposure time .
finally , the spectrum is scaled with the @xmath7-band magnitude ( section [ sec : photometry ] ) .
lr h@xmath17 @xmath184861 & 9.42@xmath19 0.68 + @xmath20@xmath21 $ ] @xmath18 5007 & 1.97 @xmath19 0.61 + @xmath20@xmath21 $ ] @xmath18 6548 & 2.32 @xmath19 0.66 + h@xmath22 @xmath18 6563 & 33.1 @xmath19 0.9 + @xmath20@xmath21 $ ] @xmath18 6584 & 7.00@xmath19 1.27 + @xmath20@xmath21 $ ] @xmath18 6717 & 4.69 @xmath19 0.52 + @xmath20@xmath21 $ ] @xmath18 6731 & 3.18 @xmath19 0.55 + [ tab : hiiregion ] lllrrrrrl 2008 november 22 & 54792.5 & 317.9 & & @xmath622.4 & @xmath23 ( 22.7 @xmath19 0.3 ) & 22.4 @xmath19 0.5 & @xmath24 & subaru + 2009 january 6 & 54837.5 & 362.9 & @xmath622.8 & @xmath622.2 & @xmath25 ( 23.2 @xmath19 0.4 ) & @xmath622.1 & @xmath26 & subaru + 2009 february 19 & 54881.2 & 406.6 & & @xmath623.6 & ( 23.9 @xmath19 0.4 ) & @xmath622.9 & & vlt + [ tab : phot ] there are very strong narrow emission lines from the region at the position of the sn . because of the faintness of the sn at nebular phases , the flux of the narrow emission lines can be easily measured .
table [ tab : hiiregion ] shows measured flux of the emission lines .
extinction is _ not _ corrected for . from the line ratios
, we can estimate gas metallicity at the site of sn 2008d .
@xcite presented metallicity measurements using the data taken soon after the explosion .
we use n2 and o3n2 indexes to estimate the metallicity @xcite : n2 @xmath27 log ( [ ] @xmath186584/ h@xmath22 ) and o3n2 @xmath27 log ( ( [ ] @xmath185007/ h@xmath17)/ ( [ ] @xmath186584/ h@xmath22 ) ) .
note that these indexes are almost independent on the extinction because of the vicinity of the lines ( [ ] and h@xmath22 , [ ] and h@xmath17 ) .
@xcite give the following relations calibrated with the metallicity estimated by @xmath28 method : 12 + log(o / h ) = 9.37 + 2.03@xmath29n2 + 1.26@xmath29n2@xmath30 + 0.32@xmath29n2@xmath31 and 8.73 @xmath32 0.32@xmath29o3n2 , for n2 and o3n2 indexes , respectively .
using these relations , metallicity is estimated to be 12+log(o / h ) = @xmath33 ( n2 ) and @xmath34 ( o3n2 ) .
the error above includes the 1@xmath35 dispersion in the relation ( 0.18 and 0.14 for n2 and o3n2 , respectively ) .
these are consistent with the solar metallicity ( 12+log(o / h)=8.66 , @xcite ) within the error , but larger than those by @xcite by @xmath360.2dex ( with the same indexes ) .
the host galaxies of the grb - associated sne ( grb 030329/sn 2003dh , grb 031203/sn 2003lw , xrf 020903 , xrf 060218/sn 2006aj ) have 12 + log(o / h ) @xmath37 in o3n2-@xmath28 scale @xcite .
thus , the metallicity at the site of sn 2008d is higher than that in those environments , as pointed by @xcite .
light curves of sn 2008d .
the red horizontal line shows estimated brightness of the region in @xmath7 band .
[ fig : lc ] ] photometric observations in @xmath38 were performed on 2008 november 22 ( @xmath39 days ) , 2009 january 6 ( @xmath8 days , figure [ fig : image ] ) , and february 19 ( @xmath40 days ) .
we detected a point source at the sn position in the @xmath7 band at three epochs and the @xmath41 band at one epoch .
a log of the observations is shown in table [ tab : phot ] .
we performed psf photometry for the @xmath7- and @xmath41-band images , where a source is detected .
for the images in which the sn was not detected , limiting magnitudes were derived using the local sky noise around the sn position , and further checked using artificial sources . since the locally strong emission of the region is present at the sn position , our @xmath7-band photometry overestimates the sn brightness .
we have corrected for this effect by estimating the contribution of the narrow lines using the observed spectrum at @xmath9 days ( @xmath42 mag , the red horizontal line in figure [ fig : lc ] ) .
the error in the sn magnitudes at @xmath39 and @xmath43 days does not include the uncertainty caused by the contamination from galaxy light .
for the data at @xmath40 days , we refrain from measuring the sn component because of the large uncertainty in the total flux as well as in the galaxy contribution
. magnitudes at early phases ( @xmath44 days , t. minezaki , in preparation ) while they are estimated using the bolometric correction for sn 2002ap at late phases @xcite .
the solid lines in blue and green show the results of radiative transfer calculation by @xcite .
the gray line shows the total energy release from 0.07 @xmath2 of @xmath1ni .
[ fig : lcbol ] ]
figure [ fig : lc ] shows the absolute magnitudes of sn 2008d . for the reddening , @xmath45 mag is assumed @xcite , and the reddening law of @xcite is adopted . for the distance to the sn , @xmath46 is assumed @xcite . the data at early phases ( @xmath47 days ) are taken from t. minezaki et al .
( in preparation ) .
we estimate the bolometric luminosity of sn 2008d at nebular phases . since multicolor observations are not available at nebular phases , we use the bolometric correction of the well - observed sn 2002ap @xcite .
we define the bolometric correction as bc @xmath48 , where @xmath49 is the bolometric magnitude constructed from @xmath50 magnitudes , and @xmath51 is the reddening corrected , absolute @xmath7-band magnitude ( as plotted in figure [ fig : lc ] ) .
the bc of sn 2008d is found to be consistent with that of sn 2002ap within 0.3 mag at early phases ( @xmath52 mag ) .
the bc of sn 2002ap at nebular phases are applied to sn 2008d by interpolating the epoch .
the estimated bc for sn 2008d is @xmath53 and @xmath54 mag at @xmath39 and 362.9 days , respectively . for the bolometric magnitudes ,
a systematic uncertainty of 0.3 mag has been added .
the bolometric light curve is shown in figure [ fig : lcbol ] .
the models by ( * ? ? ?
* blue and green lines ) , with an ejected @xmath1ni mass of 0.07 @xmath2 , are shown for comparison .
these models reproduce the bolometric light curve at early phases .
the bolometric magnitudes at nebular phases are consistent with the prediction of these models .
ejecta mass and kinetic energy are different in two models [ ( @xmath55/@xmath2 , @xmath56/@xmath57erg ) = ( 4.4 , 3.7 ) and ( 6.2 , 8.4 ) for model he6 and he8 , respectively ] .
however , since the difference in the predicted luminosity is very small , our observations at nebular phases do not discriminate between these two models . as light curves are not available ) .
the spectra are normalized to the peak flux of the [ ] @xmath58 line of sn 2008d .
the spectra are shifted by 16 , 12 , 8 , 4 , and 0 from top to bottom for clarity .
[ fig : spec ] ]
figure [ fig : spec ] shows the nebular spectrum of sn 2008d ( blue line ) taken at @xmath9 days after the explosion . the spectrum is binned into 10 , similar to the wavelength resolution at @xmath59 .
it is compared with nebular spectrum of sne 1998bw ( broad ic , * ? ? ?
* ) , 2002ap ( broad ic , * ? ? ?
* ) , 2005kl ( ic , * ? ? ?
* ) , and 2004gq ( ib , * ? ? ?
the spectrum of sn 2008d clearly shows emission lines of [ ] @xmath60 and [ ] @xmath61 .
these lines are commonly seen in the nebular spectra of other type ib / c sne .
the ir triplet is marginally detected around 8600 .
narrow lines at 4800 - 5000 and 6500 - 6900 originate in the region of the host galaxy . line profile of sn 2008d and possible contamination by the narrow [ ] line from the region . _ top _ : line profile extracted as in section [ sec : spectroscopy ] without correction of a possible narrow [ ] line from the region . _
middle _ and _ bottom _ : line profiles corrected for a possible narrow [ ] line .
the strength of the narrow [ ] line is assumed to be 1/30 and 1/15 of the h@xmath22 line , respectively .
the gray lines show unbinned spectra while the black lines show the spectra binned into 10 .
[ fig : veloi ] ] the [ ] line of sn 2008d shows a double - peaked profile .
however , given the strong narrow emission lines from the region , the broad [ ] line of the sn could possibly be contaminated by the narrow [ ] line from the region . since the strength of the narrow [ ] line is @xmath62 of h@xmath22 around the sn position ,
we subtract the scaled , gaussian - fitted h@xmath22 profile at the position of [ ] . figure [ fig : veloi ] shows the original profile ( no correction , _ top _ ) compared with the profiles corrected with 1/30 and 1/15 of the h@xmath22 line ( _ middle _ and _ bottom _ , respectively ) .
although the peak at @xmath63 km s@xmath12 in the original spectrum may be due to the contamination of the narrow emission line , the broad , double - peak profile is not affected by the contamination .
the [ ] line shows a double - peaked profile , while such clear two peaks are not seen in the [ ] line .
although these properties were also seen in the spectrum taken at @xmath64 days @xcite , transparency of the ejecta was not sure at such a transition epoch from photospheric to nebular phases .
in fact , evolution of the profile from @xmath64 days is not significant .
possible change can be seen in the redder peak of the [ ] line . at @xmath64 days
it is clearly located at @xmath63 km s@xmath12 , but the position can be redder at @xmath9 days , depending on the contamination of narrow [ ] line from the region . the [ ] line profile is compared with other sne in the left panel of figure [ fig : line ] .
double - peaked profiles are also seen in other type ib / c sne @xcite .
the fraction of sne showing a double - peaked [ ] line is roughly @xmath65 @xcite .
since our observation is late enough for the ejecta to be optically thin , the double - peaked profile of the [ ] line is unlikely to be caused by optical depth effects @xcite .
it is also unlikely to be caused by a combination of the two [ ] lines ( @xmath66 6300 and 6364 ) since the strength of the two peaks is comparable ( the strength ratio is 3:1 in optically thin limit , @xcite ) . in addition , if the double - peaked [ ] profile were caused by asphericity in the distribution of the heating source ( @xmath1ni ) , the [ ] line would also show a double - peaked profile .
thus , the profile of [ ] line reflects the distribution of excited .
the [ ] and [ ] lines may arise from different sites @xcite . [
cols="^,^ " , ] from the pole ) .
[ fig : model ] ] a simple explanation for the double - peaked [ ] profile is that oxygen has a torus - like distribution and our line of sight is near the plane of the torus ( @xcite , but see also discussion by @xcite ) .
the torus - like distribution of the presupernova elements , such as oxygen and carbon , can be realized in a bipolar explosion ( see figure [ fig : model ] for a schematic picture ) , where more energy is deposited in the polar region , and nucleosynthesis takes place there ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
a bipolar explosion has also been suggested for sn 2008d by polarimetric observations @xcite .
nebular emission profiles from bipolar models have been examined by @xcite , and compared to nebular spectra of various sne ib / c @xcite . in the right - hand panel of figure [ fig : line ] ,
the profiles calculated with a bipolar explosion model bp8 ( highly aspherical model ) by @xcite are shown for comparison .
the kinetic energy of the model is @xmath67 erg ( @xmath68 defined in @xcite ) , which is smaller than that estimated for sn
2008d ( e.g. , * ? ? ?
this is because the models of @xcite include only the c+o core . in sn 2008d , and in the models of @xcite , the he layer
is present , and it contains the rest of the kinetic energy . the line profile of sn 2008d is explained most nicely by a viewing angle of @xmath69 .
the double - peaked profile is seen only when the viewing angle is @xmath3 .
if the asphericity is smaller , this dividing angle is larger .
since the viewing angle and degree of asphericity have a degenerate effect , we can not argue that a large asphericity as in model bp8 is preferred for sn 2008d .
however , it is true irrespective of the models that our line of sight should be @xmath3 from the polar direction .
if the angle is smaller , the [ ] line would show a single - peaked profile ( see sne 1998bw , 2004gq and 2002ap and models in figure [ fig : line ] ) .
we have presented spectroscopic and photometric observations of sn 2008d at nebular phases .
flux measurements of the narrow emission lines from the region at the site of the sn and the estimates of the local metallicity were also given .
in photometric observations , the sn was detected in @xmath7 and @xmath41 .
the brightness at nebular phases is consistent with the prediction of explosion models with an ejected @xmath1ni mass of 0.07 @xmath2 , which explain the light curve at early phases . in the nebular spectrum taken at a sufficiently late phase ,
the [ ] line clearly shows a double - peaked profile , while such a profile is not seen in the [ ] line .
the evolution of the profile from @xmath70 days @xcite turned out not to be significant .
the double - peaked profile can not be explained by a spherically symmetric explosion , and it strongly indicates that sn 2008d is an aspherical explosion . the profile can be explained by a torus - like distribution of excited viewed from the side ( figure [ fig : model ] ) .
our line of sight is @xmath3 from the polar direction , irrespective of the degree of asphericity .
we are grateful to the staff of the subaru telescope for their kind support .
m.t . would like to thank yusei koyama for helpful discussion , and the anonymous referee for useful comments .
is supported by the japan society for the promotion of science research fellowship for young scientists .
this research has been supported in part by world premier international research center initiative , mext , japan , and by the grant - in - aid for scientific research of the jsps ( 18104003 , 18540231 , 20540226 , 20840007 ) and mext ( 19047004 , 20040004 ) . | we present optical spectroscopic and photometric observations of supernova ( sn ) 2008d , associated with the luminous x - ray transient 080109 , at @xmath0 days after the explosion ( nebular phases ) .
we also give flux measurements of emission lines from the region at the site of the sn , and estimates of the local metallicity .
the brightness of the sn at nebular phases is consistent with the prediction of the explosion models with an ejected @xmath1ni mass of 0.07 @xmath2 , which explains the light curve at early phases .
the [ ] line in the nebular spectrum shows a double - peaked profile while the [ ] line does not .
the double - peaked [ ] profile strongly indicates that sn 2008d is an aspherical explosion .
the profile can be explained by a torus - like distribution of oxygen viewed from near the plane of the torus .
we suggest that sn 2008d is a side - viewed , bipolar explosion with a viewing angle of @xmath3 from the polar direction . |
bl lac objects are active galactic nuclei ( agn ) , characterized by strong and variable polarization , rapid variability in luminosity , a featureless spectrum and weak optical line emission .
the radio emission associated with bl lac objects is synchrotron emission , which can be linearly polarized up to about 75% in the optically thin ( jet ) region , and up to 1015% in the optically thick ( core ) region ( pacholczyk 1970 ) .
vlbi polarization observations of bl lac objects have shown a tendency for the polarization * e * vectors in the parsec - scale jets to be aligned with the local jet direction , which implies that the corresponding * b * field is transverse to the jet , because the jet is optically thin ( gabuzda , pushkarev & cawthorne 2000 ) .
although in the past , the dominance of the transverse * b * field component was suggested to be the consequence of a ` shock model ' where a series of relativistic shocks compress and enhance the transverse * b * field component ( laing 1980 ; hughes , aller & aller 1989 ) , this seems an improbable explanation for the transverse fields detected in extended regions in the jets of some sources . instead , a helical * b * field associated with the jet , with the toroidal component dominating over the longitudinal component , would be a more plausible explanation ( lyutikov , pariev & gabuzda 2005 ) .
in fact , systematic gradients in the faraday rotation have been observed across the parsec - scale jets of a number of agn , interpreted as reflecting the systematic change in the line - of - sight component of a toroidal or helical jet * b * field across the jet ( asada et al .
2002 ; gabuzda , murray , cronin 2004 ; zavala & taylor 2005 ; gabuzda et al .
2008 ; mahmud & gabuzda 2008 ; asada et al .
2008a , b ) ; such fields would come about in a natural way as a result of the `` winding up '' of an initial `` seed '' field by the rotation of the central accreting objects ( e.g. nakamura , uchida & hirose 2001 : lovelace et al .
2002 ) .
faraday rotation studies are crucial in determining the intrinsic * b * field geometries associated with the jets .
faraday rotation of the plane of linear polarization occurs during the passage of an electromagnetic wave through a region with free electrons and a magnetic field with a non - zero component along the line - of - sight .
the amount of rotation is proportional to the integral of the density of free electrons @xmath0 multiplied by the line - of - sight b field @xmath1 , the square of the observing wavelength @xmath2 , and various physical constants ; the coefficient of @xmath2 is called the rotation measure ( rm ) : @xmath3 the intrinsic polarization angle can be obtained from the following equation : @xmath4 where @xmath5 is the observed polarization angle and @xmath6 is the intrinsic polarization angle observed if no rotation occurred ( burn 1966 ) .
simultaneous multifrequency observations thus allow the determination of the rm , as well as identifying the intrinsic polarization angles @xmath6 .
b1803 + 784 has been studied using vlbi for nearly three decades .
the predominant jet direction in centimetre - wavelength images is toward the west .
the dominant jet * b * field is perpendicular to the local jet direction essentially throughout the jet , from distances of less than 1 mas from the vlbi core ( gabuzda 1999 , lister 2001 ) to tens of mas from the core ( gabuzda & chernetskii 2003 ; hallahan & gabuzda 2009 ) ; further , the * b * field remains orthogonal even in extended regions and in the presence of appreciable bending of the jet .
therefore , it seems most likely that this transverse jet * b * field primarily represents the toroidal component of the intrinsic * b * field of the jet , rather than a series of transverse shocks ( of course , this does not rule out the possibility that some individual compact features may be shocks ) .
we have detected a transverse rm gradient across the vlbi jet ; although it is difficult to prove conclusively , combined with the observation of orthogonal * b * fields throughout the jet , even in the presence of appreciable bending , this provides direct evidence that the jet has a helical * b * field .
comparison of the gradients observed for several different epochs shows that the direction of the gradient changed sometime between june 2000 and august 2002 .
we discuss the data demonstrating this unexpected change , as well as possible origins of the observed reversal of the rm gradient .
we consider here polarization data for b1803 + 784 obtained using the ten 25-m radio telescopes of the very long baseline array ( vlba ) at four different epochs : 6 april 1997 ( gabuzda & chernetskii 2003 ) , 27 june 2000 ( zavala & taylor 2003 ) , 24 august 2002 and 22 august 2003 .
table [ tab : observations ] lists the observing frequencies for each epoch .
the observations for 24 august 2002 and 22 august 2003 were obtained as part of a multi - frequency polarization study of about three dozen bl lac objects . in all cases the sources
were observed in a ` snap - shot ' mode with 810 scans of each object spread out over the time .
the preliminary calibration , d - term calibration and polarization calibration were all done in aips using standard techniques . for more detailed calibration information for the 1997 data , see gabuzda & chernetskii ( 2003 ) and reynolds , cawthorne & gabuzda ( 2001 ) , and for the june 2000 data see zavala & taylor ( 2003 ) .
note that gabuzda & chernetskii ( 2003 ) presented results at 22 ghz in addition to the 15 , 8.4 and 5 ghz results considered here ; we did not include the 22 ghz data in our faraday rotation analysis because this substantially restricted the region in the jet where polarization was reliably detected at all the frequencies used . in regions
where polarization was detected at 22 , 15 , 8.4 and 5 ghz , the derived rotation measures are consistent with the three - frequency results presented here .
the instrumental polarizations ( d - terms ) were determined using the aips task lpcal , solving simultaneously for the source polarization in individual vlbi components .
the sources used for this purpose were 1308 + 326 ( august 2002 ) and 1156 + 295 ( august 2003 ) . the electric vector polarization angle ( evpa )
calibration was done using integrated polarization observations of a bright , compact source , obtained with the very large array ( vla ) , by forcing the evpa for the total vlbi polarization of the source to match the evpa for the integrated polarization ( www.aoc.nrao.edu/ smyers / calibration/ ) .
the sources used for the evpa calibration were 1749 + 096 ( observed with the vla on 10 august 2002 ) and 2200 + 420 ( observed with the vla on 18 august 2002 and 21 august 2003 ) .
the sources were observed with the vla at frequencies 5 , 8.5 , 22 and 43 ghz .
we found the obtained values to be consistent with the ( faraday rotation ) linear @xmath2 law and were thus able to interpolate from the graph the corresponding values for our non - standard frequencies .
we refined the calibration by checking for self - consistency between the various frequencies , and also with the other evpa calibration results for vlba experiments before and after ours that were calibrated using the same reference antenna ( the evpa corrections for a given reference antenna are typically stable to within @xmath75@xmath8 over several years ; see reynolds et al .
( 2001 ) , http://www.physics.purdue.edu/astro/mojave/sourcepages ) .
table [ tab : corrections ] shows the final applied evpa corrections .
the august 2003 data include several frequencies that differ from the standard 5 , 8 and 15 ghz frequencies for the vlba .
although system temperatures and gain curves were provided for all six frequencies observed at this epoch , there may be some concern that the accuracy of the overall flux calibration for these slightly ` non - standard ' frequencies could be slightly lower than for the standard vlba frequencies .
accordingly , we determined the spectra of various optically thin regions in the jet , to verify that there was no evidence for any of the frequencies deviating from the behaviour shown by the others . before determining the spectra , we aligned the images at the various frequencies using the algorithm of croke & gabuzda ( 2008 ) . a typical example
, corresponding to a position 3.4 mas west and 0.3 mas north of the vlbi core , is shown in fig .
[ fig : check_calibration ] ; we can see that the observed 4.615.3 ghz fluxes are all consistent with a power - law within the errors , corresponding to a `` normal '' optically thin spectral index of @xmath9 .
thus , we find no evidence for inaccuracy of the overall flux calibration at any of our six frequencies . [
tab : observations ] ll epoch & frequency ( ghz ) + 6 april 1997 & 4.990 , 8.415 , 15.285 , 22.221@xmath10 + 11 june 2000 & 8.114 , 8.209 , 8.369 , 8.594 , 12.115 , 12.591 , 15.165@xmath11 + 24 august 2002 & 15.286 , 22.234 , 43.136 + 22 august 2003 & 4.612 , 5.092 , 7.916 , 8.883 , 12.939 , 15.383 + + + [ tab : corrections ] [ cols="<,^,^,^ " , ] we made maps of the distribution of the total intensity @xmath12 and stokes parameters @xmath13 and @xmath14 at the different frequencies with matched resolutions corresponding to the lowest- or middle - frequency beam .
the distributions of the polarized flux ( @xmath15 ) and polarization angle ( @xmath16 ) were obtained from the @xmath13 and @xmath14 maps using the aips task comb .
we then constructed maps of the rm using a modified version of the aips task rm that can construct rm maps using up to 10 frequencies , obtained from r. zavala ( we will be happy to provide this code to anyone who is interested ) , after subtracting the effect of the integrated rm ( presumed to arise in our galaxy ) from the observed polarization angles ( the integrated rm value for b1803 + 784 was @xmath17 @xmath18 , determined by pushkarev 2001 ) , so that any residual faraday rotation was due to only the thermal plasma in the vicinity of the agn .
the rms and associated ( @xmath19 ) uncertainties were determined in a similar manner to that used by gabuzda , murray & cronin ( 2004 ) , taking into account the errors in the angle calibration ( 3@xmath8 ) , as well as the error in the polarization angles ( which were taken to be the rms deviation from the mean value of 3 x 3 pixel ( 0.3 x 0.3 mas@xmath20 ) area at the corresponding location ) .
[ tab : maps ] lll epoch & peak ( jy / beam ) @xmath10 & bottom contour ( mjy / beam ) @xmath10 + 24 august 2002 & 1.7 , 1.6 , 0.9 & 0.2 , 0.4 , 4.5 + 22 august 2003 & 1.3 , 1.2 , 1.3 , 1.3 , 1.3 , 1.2 & 1.6 , 1.5 , 1.6 , 1.6 , 3.3 , 5.8 + +
the total intensity ( @xmath12 ) , linear polarization ( @xmath21 ) and rm maps for epoch 11 june 2000 are presented by zavala & taylor ( 2003 ) and @xmath12 and @xmath21 maps for 6 april 1997 by gabuzda & chernetskii ( 2003 ) ; we present the rm map for the latter epoch below .
an angular distance of 1 mas corresponds to a linear distance of 7.06 pc at the redshift ( 0.68 ) of b1803 + 784 ( http://www.physics.purdue.edu/mojave/sourcepages/1803+784.shtml ) .
[ fig : slava_maps ] and [ fig : my_pol_maps ] show the observed vlbi total intensity and linear polarization structures for august 2002 and august 2003 .
the peaks and bottom contours of these maps are given in table [ tab : maps ] .
the images in figs .
[ fig : apr1997jet ] , [ fig : aug2002jet ] and [ fig : aug2003jet ] show the vlbi total intensity contours for 6 april 1997 , 24 august 2002 and 22 august 2003 with their parsec - scale rm distributions superimposed . in all the maps presented here , the contours increase in steps of a factor of two , and the beam is shown in the lower left corner of the image .
the polarization maps for all four epochs are consistent with a predominantly transverse magnetic - field structure .
transverse rm gradients are visible across the vlbi jet of b1803 + 784 in each of the three epochs shown in figs.[fig : apr1997jet][fig : aug2003jet ] , on scales of @xmath22 mas ; we are also tentatively able to follow the rm gradient further from the core in fig .
[ fig : aug2003jet ] .
this supports the hypothesis that this jet has a helical magnetic field , consistent with the observed transverse magnetic - field structure .
the arrows show the directions of the rm gradients in the corresponding regions ; in other words , the direction in which the value of the rm increases ( from more negative to less negative , negative to positive , or less positive to more positive , as the case may be ) . the accompanying panels show plots of polarization angle ( @xmath19 ) vs. wavelength squared ( @xmath23 ) for the indicated regions , as well as slices through the rm distributions between the indicated points .
the uncertainties in the polarization angles are also shown in the plots .
[ fig : trans_dist ] shows plots of the observed rm as a function of the transverse distance from the central spine of the jet .
the errors in the rms in these plots were estimated in the same way as for the @xmath19 values , by calculating the mean rm within the corresponding @xmath24 pixel area in the map and assigning the rms deviation about this mean as the rm error .
we consider this approach to be most reasonable , since possible evpa calibration errors will affect the fitted rm values and the uncertainties derived from the fits , but not the inferred rm gradients , as is shown in the appendix .
both these plots and the rm slices in figs .
[ fig : apr1997jet ] [ fig : aug2003jet ] clearly demonstrate the systematic , monotonic nature of the observed rm gradients .
the rm map of zavala & taylor ( 2003 ) for june 2000 also shows a clear transverse gradient , with a negative rm along the northern edge of the jet and less negative or positive rm along the southern edge , similar to the rm map observed in august 2002 ( fig .
[ fig : aug2002jet ] ) .
an important feature that emerges in a comparison of the various rm maps is the reversal in the direction of the rm gradient in the jet of b1803 + 784 between june 2000 and august 2002 .
the rm values increase toward the southern edge of the jet in the april 1997 and june 2000 ( zavala & taylor 2003 ) rm maps , but toward the northern edge of the jet in the august 2002 and august 2003 rm maps .
we also observe gradients in the core region for all 3 of our epochs , shown by the arrows in the core region in figs .
[ fig : apr1997jet ] , [ fig : aug2002jet ] and [ fig : aug2003jet ] , all of which are in the same direction ( i.e. , the `` flip '' of the gradient occurred only in the jet region ) .
there are hints of a rm gradient with the same direction in the core in the map presented by zavala & taylor ( 2003 ) , but the range for the colour scale of that map makes it difficult to be sure of this .
note that the observed `` core '' emission is actually a mixture of emission from the core and inner jet , so it is not surprising that transverse rm gradients should be detected in the core region , since the associated helical field is present on a wide range of scales , down to the accretion disk region .
the origin for the observed `` flip '' in the direction of the transverse rm gradient in the jet of b1803 + 784 between june 2000 and august 2002 is not clear .
no radical change in the intensity or polarization structure of the vlbi jet accompanied the observed rm - gradient flip .
we qualitatively consider here several possible scenarios that could potentially give rise to this behaviour .
although it is not possible at this stage to distinguish between these scenarios , our goal here is simply to point out possible explanations , which it may be possible to test with further observations or numerical simulations . in this discussion
, we will suppose that the rm gradient is associated with a helical jet * b * field that is due to the rotation of the central black hole . in this case , the direction of the observed transverse rm gradient is determined by the direction of rotation of the black hole and the `` polarity '' of the poloidal field component , i.e , north or south / outward or inward relative to the black hole .
* reversal of the `` pole '' of the black hole facing the earth . *
it is obviously not reasonable to suppose that the observed reversal in the rm gradient is due to a change in the direction of rotation of the central black hole .
however , one way to retain a transverse rm gradient in a helical magnetic field model but reverse the direction of this gradient , is if the direction of rotation of the central black hole ( i.e. the direction in which the field threading the accretion disc is `` wound up '' ) remains constant , but the `` pole '' of the black hole facing the earth changes from north to south , or vice versa ( fig .
[ fig : north_south_rev ] ) . to our knowledge , it is currently not known whether such polarity reversals are possible for agn jets , or whether they could occur on a time scale as short as a few years .
numerical simulations of the generation and propagation of helical magnetic fields for various initial seed - field configurations could potentially answer these questions . * torsional oscillations of the jet . *
another way in which the direction of the azimuthal component of a helical * b * field could change is as a result of torsional oscillations of the jet , such as those that have been proposed by bisnovatyi - kogan ( 2007 ) .
these torsional oscillations , which may stabilize the jets , essentially correspond to a change in the direction of rotation of the jet , and could cause reversals of the azimuthal * b * field component from time to time , probably periodically or quasi - periodically .
of course , we have only observed a single reversal in the direction of the rm gradient , and so have no information at this stage about whether this phenomena is periodic .
therefore , this model remains a possible explanation for the observed rm - gradient reversal , but we currently have no firm evidence in support of this model .
if future multi - frequency vlba polarization observations show the presence of periodic reversals of the direction of a transverse rm gradient across the jet , this will provide support for this type of model .
* a `` nested - helix '' magnetic - field structure . * another intriguing possibility is related to magnetic - tower - type scenarios for the launching of jets ( lynden - bell 1996 ) . in this picture ,
magnetic field lines anchored in the inner accretion disc are carried outward along with the jet flow and return in outer layers of the jet , closing through the outer part of the accretion disc .
combined with the effect of the differential rotation of the accretion disc , this gives rise to a `` nested helical field '' structure , with an `` inner '' helical field displaying a helicity corresponding to the direction of rotation of the accretion disc and the magnetic polarity of the jet base and an `` outer '' helical field displaying the opposite helicity ( fig .
[ fig : nested_hlx ] ) .
the pitch angles of the inner and outer helical fields may be different , most likely with the inner field having a lower pitch angle ( being more tightly wound ) .
the net rm gradient we observe will include contributions from both of these fields , corresponding to the volumes of both of these regions that lie along the line of sight between the emission region and the observer .
the directions of the rm gradients associated with each region will be opposite , and the net observed rm gradient will be determined by whether the inner or outer region makes the larger contribution to the net observed rm . in turn , this will be determined by the distribution of the electron density and magnetic field along the jet and with distance from the jet axis .
the direction of the observed rm gradient could `` flip '' if , for example , the outer region of helical field usually dominates the observed rm distribution , but the inner region of helical field temporarily became dominant due to an increase in the electron density or magnetic - field strength in the inner region of helical field .
such variations in electron density and/or magnetic - field strength could come about due to variations in activity of the central engine and the ejection of jet material . in this case , the resulting reversals in the direction of the net observed rm gradient would most likely not occur periodically , and may follow activity of the vlbi core .
if future observations provide evidence for a connection between epochs when the direction of the transverse rm gradient reverses and periods of enhanced core activity , this will provide support for this type of scenario . * changes in the distribution of thermal electrons surrounding the jet .
* finally , we should consider whether the observed change in the direction of the transverse rm gradient could be due purely to changes in the distribution of thermal gas in the vicinity of the jet , unrelated to the presence of a helical jet magnetic field .
could the transverse gradient in the rm be due , for example , to a gradient in the thermal electron density across the jet , with the change in the direction of the gradient corresponding to a change in the side of the jet where the electron density was highest ?
in this scenario , the changes in the observed rm distribution would essentially be due to `` patchiness '' in a moving intervening faraday screen , and so would probably occur fairly randomly with time .
we find this explanation unlikely for several reasons .
the first is that our observations and the observations of zavala & taylor ( 2003 ) consistently show the presence of transverse rm gradients over roughly six years , whereas we would expect the rm pattern produced by a moving `` patchy '' screen not to be as consistent .
further , in our 1997 rm map ( fig .
[ fig : apr1997jet ] ) , the rm on the northern side of the jet is negative , whereas the rm on the southern side of the jet is positive : a change of the sign of the rm from one side of the jet to the other is difficult to explain as the result of a patchy thermal - electron distribution , but is quite natural if the observed rm gradient is due to a helical jet magnetic field . with regard to the observed changes in the rm distribution , a comparison of our 1997 and 2003 rm maps ( figs .
[ fig : apr1997jet ] and [ fig : aug2003jet ] ) shows that the rm on the northern side of the jet is negative in 1997 , but positive in 2003 . here ,
also , a change in the thermal - electron distribution is not sufficient to explain these observations : they require that the direction of the line - of - sight component of the magnetic field in the region of faraday rotation has changed between these two epochs .
since a change in the line - of - sight magnetic field is required , this makes it natural to consider scenarios involving changes in a helical jet magnetic field structure .
finally , we note that another possible way to explain time - variable faraday rotation is if compact jet components `` illuminate '' different regions in a static , inhomogeneous foreground faraday screen as they move behind it , as was suggested by zavala & taylor ( 2001 ) for the core rm variations in 3c279 .
however , this is likewise an unlikely explanation for the rm changes observed for b1803 + 784 , because it is difficult to imagine how this effect could give rise to changes in the rm pattern _ across _ the jet .
in addition , the apparent speeds of vlbi components near the region where we have detected the transverse rm gradients in b1803 + 784 are low : an essentially quasi - stationary component has been observed roughly 1.5 mas from the core for nearly three decades , and no features detected closer than a few mas from the core show clear motions ( kellermann et al .
2004 ; gabuzda & chernetskii 2003 ; http://www.physics.purdue.edu/mojave/sepavstime/b1803 + 784_sepvstime.gif ) .
this likewise appears to make it unlikely that the scenario envisaged for 3c279 is operating in this region of the b1803 + 784 jet , since none of the observed compact vlbi features display appreciable motions .
it is interesting to consider the possible motion or propagation of the transverse rm gradient pattern . to compare the distances from the core where the transverse rm gradient is most clearly visible
, we must construct rm maps for the different epochs using the same restoring beam .
we have used the beam in fig .
[ fig : apr1997jet ] for this purpose , and the rm maps for august 2002 and august 2003 constructed using this beam are shown in fig .
[ fig : same_beam ] .
we first note that the transverse gradients we initially identified based on the images with their intrinsic beams are still visible in each of these matched - resolution maps .
this is particularly important for the august 2002 rm image , which was made from data obtained at 1543 ghz , rather than 515 ghz , and so had an intrinsically appreciably higher resolution . as in fig .
[ fig : apr1997jet ] , the transverse rm gradient is visible essentially all along the vlbi structure , out to nearly 3 mas from the core .
the distances from the core where the transverse rm gradient is most clearly visible in the remaining two matched resolution images are roughly @xmath25 mas in april 1997 and @xmath26 mas in august 2003 ; in the june 2000 rm map of zavala & taylor ( 2003 ) , the rm gradient is most prominent @xmath27 mas from the core .
( note that we do not have access to the actual june 2000 rm map of zavala & taylor ( 2003 ) in electronic form , and so our ability to analyze this image and compare it with the other images considered is quite limited . )
the april 1997 , june 2000 and august 2003 rm maps all have similar intrinsic resolutions , and so the positions where the transverse rm gradients are most clearly visible can be meaningfully compared .
however , we see no clear evidence for systematic movement of the observed rm patterns either toward or away from the vlbi core over the time covered by these epochs : the distances from the core are @xmath25 mas in april 1997 , @xmath27 mas in june 2000 , and @xmath26 mas in august 2003 .
this is not entirely surprising , since the separation between our epochs may not be well matched to the time scales on which the rm patterns evolve and/or propagate .
for example , if the rm pattern moves with a proper motion of several tenths of a mas per year ( typical of superluminal motions in agn with redshifts similar to b1803 + 784 ) , it could move and evolve substantially over the three - year intervals from april 1997 to june 2000 and from june 2000 to august 2003 , making it difficult to track the pattern . at the same time , the general region of the transverse rm gradient is also near the location of a long - lived quasi - stationary feature in the jet of b1803 + 784 , approximately 1.5 mas from the vlbi core ( e.g. kellermann et al . 2004 ; gabuzda & chernetskii 2003 ) .
depending on the origin of this quasi - stationary feature , it may be that rm patterns in this region will also display very low apparent speeds along the jet .
more frequent multi - frequency monitoring of this agn with polarization vlbi could provide additional information about the evolution of the rm structure on parsec scales , which could potentially provide firmer constraints on possible models for the origin of the observed rm gradient and its evolution .
we have detected transverse rm gradients across the vlbi jet of b1803 + 784 at three different epochs ; a transverse rm gradient is also visible in the rm distribution for this agn published by zavala & taylor ( 2003 ) .
the presence of a transverse rm gradient across the jet at four different epochs spanning about 7 years provides firm evidence for a helical * b * field associated with this jet .
unexpectedly , a comparison of the rm gradients for these 4 epochs shows a clear _ reversal _ of the direction of the gradient between june 2000 and august 2002 .
this is not accompanied by any obvious change in the jet intensity or polarisation structure .
the origin for this observed `` flip '' in the direction of the transverse rm gradient in the jet of b1803 + 784 is not clear .
we have suggested several scenarios that could give potentially rise to this phenomenon : ( i ) a reversal of the direction of the poloidal component of the intrinsic magnetic field of the central black hole ; ( ii ) a reversal of the direction of the azimuthal field component associated with torsional oscillations of the jet ; and ( iii ) a change in whether the `` inner '' or `` outer '' region of helical magnetic field dominates the total observed rm in a magnetic - tower picture . although it is not possible at this stage to conclusively identify whether any of these scenarios is the origin of the observed rm - gradient reversal , we favour the last hypothesis , because it provides a relatively simple explanation for this seemingly strange event , since a `` nested helix '' structure for the jet magnetic field is a natural outcome of magnetic - tower models .
further multi - frequency polarization studies of this source are clearly crucial for our understanding of how the transverse faraday rotation gradients evolve over time and with distance from the core .
higher resolution multi - frequency vlbi polarization studies can be used to study transverse rm gradients within the observed vlbi core in more detail , as is demonstrated by our 1543 ghz rm image .
we have recently obtained further vlba polarization observations of b1803 + 784 at 543 ghz , which we hope will enable us to trace the behaviour of the rm distribution over a wider range of distances along the jet , all at a single epoch .
numerical studies to determine the feasibility of various scenarios for the reversal of the transverse rm gradient would also be of considerable interest , and are planned as part of our continuing work in this area .
the research for this publication was financially supported by a basic research grant from science foundation ireland .
the national radio astronomy observatory is operated by associated universities inc .
we thank p. cronin for his work on the rm map for april 1997 , t. v. cawthorne for useful discussions of these results , and the anonymous referee for useful comments and suggestions that have improved this paper .
asada , k. , inoue , m. , uchida , y. , kameno , s. , fujisawa , k. , iguchi , s. mutoh , m. 2002 , pasj , 54 , l39 asada k. , inoue m. , kameno s. , nagai h. 2008a , apj , 675 , 79 asada k. , inoue m. , nakamura m. , kameno s. , nagai h. 2008b , apj , 682 , 798 bisnovatyi - kogan g.s .
2007 , mnras , 376 , 1 , 457 burn b.j . 1966 ,
mnras , 133 , 67 gabuzda d. c. 1999 , new astron .
, 43 , 691 gabuzda d. c. , chernetskii v. a. 2003 , mnras , 339 , 3 , 669 gabuzda d. c. , murray . & cronin p. j. 2004 , mnras , 351 , l89 gabuzda d. c. , pushkarev .
a. b , & cawthorne t. v. 2000 , mnras , 319 , 1109 gabuzda d. c. , vitrishchak v. m. , mahmud m. & osullivan s. p. 2008 , mnras , 384 , 1003 hughes p.a .
, aller h.d . &
aller m.f .
1989 , apj , 1 , 349 , 54 laing r. a. 1980 , mnras , 193 , 439 lister m. l. 2001 , apj , 562 , 208 lovelace r. v. e. , li h. , koldoba a. v. , ustyugova g. v. & romanova m. m. 2002 , apj , 572 , 445 lynden - bell d. 1996 , mnras , 279 , 389l .
lyutikov m. , pariev v. & gabuzda d. c. 2005 , mnras , 360 , 869 mahmud m. & gabuzda d. c. 2008 , in _ extragalactic jets : theory and observations from radio to gamma - rays _
travis a. rector and david s. de young , asp conf .
, 386 , 494 nakamura m. , uchida y. & hirose s. 2001 , new astronomy , 6 , 2 , 61 pacholczyk , a. g. , 1970 , radio astrophysics , w. h. freeeman , san franciso pushkarev a. b. 2001 , astron . rep . , 45(9 ) , 667 reynolds , c. , cawthorne t.v .
& gabuzda d.c .
2001 , mnras , 327 , 4 , 1071 zavala r. t. & taylor g. b. 2003 , apj , 589 , 126 zavala r. t. & taylor g. b. 2005 , apj , 625 , l73 http://www.physics.purdue.edu/mojave/sepavstime/b1803 + 784_sepvstime.gif http://www.physics.purdue.edu/mojave/sourcepages http://www.physics.purdue.edu/mojave/sourcepages/1803+784.shtml
to illustrate why the presence of a rm gradient will not be affected by incorrect evpa calibration , we consider first the definition of the rm , measured by comparing the polarisation angles @xmath28 and @xmath29 measured at two wavelengths @xmath30 and @xmath31 : we can see through the simple rearrangement in the second equation above that , although uncertainty in the absolute evpa calibration will contribute to the absolute uncertainties of the rm values , the effects of incorrect evpa calibration essentially cancel out when calculating the rm _ gradient _ , since they will appear in the same way in the polarisation angles measured at each point in the image at the wavelength in question , e.g. , in @xmath36 and @xmath37 .
this fact must be taken into account when comparing rm values at different points in an image and their uncertainties in order to estimate the significance of observed rm gradients .
for this reason , we have estimated the uncertainties in the rm values at a given location in the map in fig . [
fig : trans_dist ] purely from the rms deviation of the rm values in a @xmath24-pixel area surrounding this location , rather than using the formal uncertainties from the fits yielding the rm values , since the latter will be affected by possible inaccuracies in the evpa calibrations . | several multi - frequency polarization studies have shown the presence of systematic faraday rotation gradients across the parsec - scale jets of active galactic nuclei ( agn ) , taken to be due to the systematic variation of the line - of - sight component of a helical magnetic ( * b * ) field across the jet .
other studies have confirmed the presence and sense of these gradients in several sources , thus providing evidence that these gradients persist over time and over large distances from the core .
however , we find surprising new evidence for a reversal in the direction of the faraday rotation gradient across the jet of b1803 + 784 , for which multi - frequency polarization observations are available at four epochs . at our three epochs and the epoch of zavala & taylor ( 2003 ) ,
we observe transverse rotation measure ( rm ) gradients across the jet , consistent with the presence of a helical magnetic field wrapped around the jet .
however , we also observe a `` flip '' in the direction of the gradient between june 2000 and august 2002 . although the origins of this phenomena are not entirely clear , possibly explanations include ( i ) the sense of rotation of the central supermassive black hole and accretion disc has remained the same , but the dominant magnetic pole facing the earth has changed from north to south ; ( ii ) a change in the direction of the azimuthal * b * field component as a result of torsional oscillations of the jet ; and ( iii ) a change in the relative contributions to the observed rotation measures of the `` inner '' and `` outer '' helical fields in a magnetic - tower model .
although we can not entirely rule out the possibility that the observed changes in the rm distribution are associated instead with changes in the thermal - electron distribution in the vicinity of the jet , we argue that this explanation is unlikely .
[ firstpage ] |
several unification schemes merging gravity and the standard model of strong and electroweak interactions predict the existence of short - range forces with coupling strength of the order of newtonian gravity @xcite .
efforts to evidence a _
fifth force _ have been envisaged regardless of any concrete unification scheme since various decades @xcite , and there are compelling reasons to improve our limits especially in the largely unexplored submillimeter range .
constraints in both coupling and range for these interactions have been obtained with various experimental setups , including the recent configurations using a disk - shaped torsional balance parallel to a rotating flat surface @xcite , or micromechanical resonators in a parallel plane geometry @xcite . due to the surge of activity in the study of casimir forces
, limits have been also given in the submicrometer range based on the level of accuracy between casimir theory and experiment .
however , unlike the case of experiments performed between bodies kept at larger distances , the use of the parallel plane geometry on such small lengthscale has been proven to be challenging in terms of parallelism @xcite , and therefore the attention has been focused on the analysis of the residuals in the casimir theory - experiment comparison involving the sphere - plane configuration .
dedicated efforts to obtain limits from sphere - plane casimir experiments have involved the use of the so - called proximity force approximation ( pfa ) @xcite , which allows to map the force @xmath1 between a sphere of radius @xmath2 and a plane located at a distance @xmath3 from the sphere into the energy per unit area @xmath4 of the parallel plate configuration , namely @xmath5 @xcite .
this approximation is believed to be valid in the limit @xmath6 and to hold with a high degree of accuracy for forces between entities concentrated on the surfaces , such as electrostatic or casimir forces between conductors @xcite .
obviously , in order to test how well pfa approximates the exact force , one needs either to compute the interaction exactly or at least to assess reliable bounds . for the electrostatic sphere - plane interaction , the exact analytical result for
the force is well - known and has a closed form @xcite , such that deviations from pfa can be readily analyzed . for the casimir sphere - plane interaction ,
the exact force has been computed only very recently , both for ideal @xcite and real metallic plates @xcite .
available analytical and numerical results seem to indicate that , at least for zero temperature and within the used plasma model , deviations from pfa applied to the sphere - plane casimir interaction are small , of the order of @xmath7 or higher , in recent casimir experiments aiming to put limits to yukawa interactions .
it has been argued in @xcite that the application of the pfa to forces acting between entities embedded in volumetric manifolds , such as gravitational forces or their putative short - range components , is in general _ invalid _ and has to be carefully scrutinized in each specific configuration .
based on this suggestion , a recent reanalysis of the pfa in the case of gravitational and yukawian forces has been discussed in @xcite .
the main conclusion of this reanalysis is that `` a confusion with different formulations of the pfa '' existed in the previous literature , and that `` care is required in the application of the pfa to gravitational forces '' .
this confusion is stated to originate from a specific form of the pfa used so far , to be contrasted with a more general formulation of the pfa . in @xcite
it is also claimed that the difference between the two pfas is negligible in the actual configuration used to give the allegedly best limits obtained in the 100 nm range @xcite . in this paper
we further discuss the meaning of the pfa in the case of volumetric forces .
we argue that the discrepancy between the two forms of the pfa is a significant source of error in the determination of bounds on parameters of yukawian forces from force residuals in casimir sphere - plane experiments performed so far that used pfa to model such non - newtonian forces .
we then show that the general form for the pfa discussed in @xcite is simply a different choice of the infinitesimal volume for integrating the force due to an extended object , and coincides with the exact result only in the case when one of the two surfaces is an infinite planar slab ( or semispace ) .
the level of approximation in using the two pfas for yukawa forces in the sphere - plane geometry is of the same order of magnitude as the casimir theory - experiment comparison , that uses pfa to compute the sphere - plane casimir force ( as already noticed in @xcite ) .
therefore , since such a comparison provides force _ residuals _ that are in turn compared against the theory of yukawa forces to obtain limits on its @xmath0 parameter space , the use of these _ subsequent _ pfa approximations of comparable level of approximation provides a possible source of systematic error , not carefully accounted for so far .
we also argue that other volumetric effects not directly related to the pfa , such as the finite size of the planar surface used in the actual experiments , may provide a source of systematic error not taken into account so far , which strongly affects the limits to power - law forces @xcite , but should not be a major source of concern on limits to yukawa forces .
we believe that , considering the various number of complications related to the sphere - plane geometry , upgraded versions of parallel plates experiments such as the ones discussed in @xcite could provide limits on yukawian and power - law forces in the submicrometer range more immune to a set of systematic errors characteristic of the sphere - plane configuration .
in order to introduce the notation and as a prelude to our discussion , we briefly summarize the results contained in @xcite .
the actual experimental configuration used in @xcite is not a parallel plate geometry , rather it is a sphere - plane geometry , and the pfa is used to map the force between a sphere and a plane @xmath8 into the energy per unit area of the parallel plate configuration @xmath9 @xmath10 where @xmath11 is the geometrical average of the principal radii of curvature of the spherical surface evaluated at its point of minimum distance from the plane . in the experiment reported in @xcite ,
the force is measured by looking at the frequency shift of a mechanical resonator , as customary in atomic force microscopy @xcite , and as first reported in the context of casimir force measurements in @xcite .
the frequency shift is proportional to the gradient of the force , and therefore @xmath12 where @xmath13 is the plane - plane pressure , and @xmath14 is the mass of the resonator .
the measure of the frequency shift can then be mapped , via use of eq .
( [ approxpfa ] ) , into the equivalent pressure exerted between two fictitious parallel plates mimicking the actual sphere - plane geometry . within the validity of eq .
( [ approxpfa ] ) , this is a valid assumption for the case of forces acting between surfaces , such as electrostatic forces between conductors or casimir forces .
a first sign of the fact that there can be issues with the pfa in dealing with volumetric forces , such as the hypothetical yukawian forces of gravitational origin , is manifested by noticing that the exact formula for the yukawa force between two infinite parallel slabs depends on the thicknesses of both slabs , which implies that the pfa formula applied to the volumetric yukawa force in the sphere - slab configuration also depends on _ both _ thicknesses ( and on the sphere radius ) .
however , the exact sphere - slab force obviously depends only on the slab thickness and on the sphere radius - it does not , and can not , depend on the thickness of the _ metaphysical _ slab introduced in the virtual mapping to the parallel geometry .
indeed , consider the yukawa potential energy for two pointlike masses @xmath15 and @xmath16 , located at positions @xmath17 and @xmath18 respectively , @xmath19 where , as usual , the strength of the yukawa interaction is parameterized in terms of newton s gravitational constant @xmath20 through a dimensionless quantity @xmath21 , and @xmath22 is its range .
assuming that the yukawa interaction is additive , once integrated over two infinite , homogeneous parallel slabs separated by a distance @xmath3 , one derives the corresponding pressure @xmath23 , @xmath24 where @xmath25 and @xmath26 indicate the thickness of each slab , @xmath27 and @xmath28 their densities . the exact yukawa interaction in the sphere - slab geometry can be readily computed , assuming additivity .
the result is @xcite @xmath29 as we mentioned above , most recent experimental works on limits to extra - gravitational forces from sphere - plane casimir measurements used the usual pfa . in this approximation ,
the yukawa force between a homogeneous sphere and an infinite homogeneous slab of thickness @xmath25 is @xmath30 in this case one needed to consider , in order to map the actual sphere - plane configuration into a parallel plate geometry , a fictitious upper plate of thickness @xmath26 large enough , _
i.e. _ much larger than the explored yukawa range ( @xmath31 ) . again , this situation may appear disturbing to whoever believes that any experiment - theory comparison should not rely on the introduction of arbitrary parameters not having a tangible , measurable counterpart in the concrete experimental setup .
clearly the pfa prediction eq .
( [ yukpfa ] ) fails to give the exact result eq .
( [ yukexact ] ) in the range of its supposed validity , @xmath6 , and it is necessary to assume @xmath32 in order for pfa to tend to the exact result . but in this limit the volumetric nature of the interaction is lost , since the atoms in the bulk " no longer contribute appreciably to the total force .
likewise , pfa fails to give the exact newtonian interaction between the sphere and the slab , even in the range of its supposed validity @xmath6 . due to
the plane is satisfied ) , these merely correspond to two different and equivalent choices for the integration volume.,title="fig:",scaledwidth=45.0% ] due to the plane is satisfied ) , these merely correspond to two different and equivalent choices for the integration volume.,title="fig:",scaledwidth=45.0% ] the authors of @xcite consider the most general formulation of pfa @xcite , that we will call exact " pfa formula ( epfa ) to distinguish it from the usual pfa approximation , described above . in the epfa
the force between two compact bodies is expressed as the sum of forces between plane parallel surface elements @xmath33 .
the @xmath34 component of the force is @xmath35 where @xmath36 is the pressure between two parallel plates at a local distance @xmath37 ( @xmath38 are the surfaces of the two bodies ) , @xmath3 is the distance between them ( smallest value of @xmath39 ) , and @xmath40 is the part of the @xmath41-plane where both surfaces are defined ( see fig .
the epfa prediction for the yukawa force between a sphere and an infinite planar slab is @xcite @xmath42 which coincides with the exact result of eq.([yukexact ] ) . in fig.2
we plot the ratio @xmath43 @xmath44 as a function of the yukawa parameter @xmath22 , for different values of the sphere radius keeping fixed @xmath45 ( left plot ) , and for different values of @xmath26 keeping fixed the sphere radius at @xmath46 m ( right plot ) .
note that @xmath43 is independent of the sphere - slab separation @xmath3 .
when @xmath47 , as surely realized in the left plot , pfa always _ overestimates _ the epfa result , i.e. @xmath48 ( similarly to how the pfa overestimates the exact casimir force in the sphere - plane geometry ) . note also that when @xmath31 the atoms in the bulk " of the two bodies do not contribute appreciably to the yukawa force , thereby making it effectively of a surface character ( i.e. , non - volumetric ) , as in the case of casimir or electrostatic forces .
instead , for values of @xmath49 ( or smaller ) the volumetric character of the yukawa interaction is manifest , and @xmath43 is no longer less than one ( right plot ) . in this case
the pfa applied to yukawa forces is invalid and in the data analysis one should at least declare the value of @xmath26 imagined for which the limits are assessed . overestimating the yukawa force leads to stronger limits for the coupling constant @xmath21 for a given @xmath22 with respect to the proper use of the epfa . considering the relatively small margins of improvement reported recently ( see for instance fig . 3 in @xcite ) a systematic shift due to the use of the pfa instead of the epfa may lead to significant changes for the exclusion region in the @xmath0 plane .
both plots show that the use of pfa instead of epfa is unreliable especially in the region near or above @xmath22=100 nm . unfortunately the region in between 100 nm and few @xmath50 m is also the one _ directly _ explored with the casimir force experiments , since actual measurements take place in this range of distance between the involved objects .
it is known that the best limits on yukawa interactions can be set for @xmath22 of the order of the actual explored distance between the two bodies @xmath51 , and the extrapolation of the measurements to smaller @xmath22 is affected by the fast growth of the bounds as @xmath52 .
the fact that the epfa eq .
( [ yukepfa ] ) gives the correct exact results for the yukawa ( and also gravitational ) force in the sphere - plane configuration is in fact a trivial consequence of the additivity of these interactions and of the translational invariance of the infinite plane surface , the shape of second surface ( in this case a sphere ) being irrelevant .
indeed , the epfa is just a _ different _ parameterization of the exact formula of addition of forces between particles . on one hand ,
the exact interaction energy between a test mass and an infinite slab ( or half - space ) depends only on the normal coordinate @xmath34 , being independent of the in - plane coordinates @xmath53 by symmetry .
hence , the potential due to the infinite slab is @xmath54 . for a body ( _ e.g.
_ a sphere ) of mass density @xmath55 , additivity implies that the total interaction energy can be obtained as @xmath56 and similarly for the force . since @xmath57 depends only on @xmath34 , it is convenient to compute the integral by adding forces at different slices at constant @xmath34 , _
i.e. _ considering infinitesimal slices in @xmath34 , then evaluating first the potential energy of a slice of the body parallel to the plane at a distance @xmath34 @xmath58 and then integrating along @xmath34 the quantity @xmath59 one obtains the exact expression for the body - plane interaction @xmath60 ( see fig .
on the other hand , the epfa states that the interaction energy between the body and the plane is obtained from slicing the body into cylinders perpendicular to the plane ( see fig . 1 right ) and integrating the cylinder - plane interaction energy along the portion of the body that faces the plane ( i.e.
, one must integrate over the surface @xmath40 on the plane that is the normal shadow of the body ) .
the potential energy of this column of the body centered around @xmath41 is @xmath61 and then integrating along @xmath62 the quantity @xmath63 one gets the epfa expression for the body - plane interaction @xmath60 ( see fig . 1 right ) .
again , since the interaction is additive and it does not depend on the @xmath62 coordinates , this integral is exactly equal to the previous one : we are simply integrating the same function @xmath64 over the body using a different parameterization of the volumetric integral .
the same holds for the force between any body ( not necessarily a sphere ) and an infinite slab ( see also @xcite ) . versus @xmath22 in the case of spheres of radius @xmath65150 , 100 , and 50@xmath50 m , in the limit of @xmath66 .
( right ) same plot but for a sphere of radius @xmath67 m and finite values of the metaphysical parameter @xmath26 representing the thickness of the upper slab introduced in @xcite for the comparison of the actual sphere - plane geometry to the parallel plate case using the pfa.,title="fig:",scaledwidth=45.0% ] versus @xmath22 in the case of spheres of radius @xmath65150 , 100 , and 50@xmath50 m , in the limit of @xmath66 .
( right ) same plot but for a sphere of radius @xmath67 m and finite values of the metaphysical parameter @xmath26 representing the thickness of the upper slab introduced in @xcite for the comparison of the actual sphere - plane geometry to the parallel plate case using the pfa.,title="fig:",scaledwidth=45.0% ] however , when none of the two bodies is an infinite slab ( e.g. , two spheres of radii @xmath68 , mass densities @xmath69 , separated by a distance @xmath3 along the @xmath34 direction ) , translational invariance along @xmath70 is obviously broken , and the epfa does not coincide with the exact formula .
the exact interaction energy @xmath71 between a source body and a test mass at position @xmath72 can be easily computed .
for instance , for the spherical source body @xmath73 depends only on @xmath74 , with the origin of coordinates at the center of the sphere . integrating @xmath71 over the volume of the second body one gets the exact result .
for example , for the two - spheres case the gravitational energy depends only on the center - to - center distance and scales as @xmath75 .
let us compare this known exact result with the epfa prediction .
one slices each body in cylindrical slabs , calculates the slab - slab interaction energy @xmath76 that depends on the local distance @xmath34 between the slabs , and finally one adds up these contributions over the shadow of one of the bodies on the other one .
it is clear that epfa can not give the exact result since @xmath77 is translational invariant but the exact @xmath73 is not , and epfa fails to predict the exact @xmath75 dependency .
therefore the epfa formula for the energy and force of additive two - body interactions is a trivial reparameterization of the exact result when one of the bodies is an infinite plane ( or slab ) . for other geometries
, epfa fails to give the correct result as a consequence of the broken translational invariance . in particular
, this is the case of a sphere above a finite size slab , like in experiments , especially those involving slabs of typical sizes comparable to those of the sphere . in the experimental configuration used in @xcite various substrates
are present on the sphere and on the slab . imagining that the layered slab is infinitely long , the yukawa potential at a distance @xmath34 from the top layer due to the slab is @xmath78 .
\label{testmass } \end{aligned}\ ] ] here @xmath79 and @xmath80 are the thickness and density of the top layer , @xmath81 and @xmath82 are the thickness and density of the middle layer , and @xmath25 and @xmath27 and the thickness and density of the lower part of the layered slab .
the last factor in eq.([testmass ] ) can be considered a sort of _ effective _ density of the planar surface , in which the various densities are weighted by their thicknesses in units of @xmath22 ( indeed yielding their arithmetic average in the case of @xmath83 ) .
for the comparison of the epfa and pfa schemes for the multilayered and corresponding homogeneous situation ( obtainable by using eqs .
( 14 ) and ( 15 ) with @xmath96 , @xmath97 , and replacing @xmath2 with @xmath98 ) .
( left ) ratio @xmath99 versus the range of the yukawian force for different values of the radius of the inner sphere .
the parameters for the layered sphere are @xmath100 nm , @xmath101 nm , @xmath28=4.1 g/@xmath102 , @xmath103=7.14 g/@xmath102 , and @xmath104=19.28 g/@xmath102 .
the parameters for the layered slab are @xmath105 m , @xmath106 nm , @xmath107 nm , @xmath27=2.33 g/@xmath102 , @xmath108=7.14 g/@xmath102 , and @xmath109=19.28 g/@xmath102 . in the evaluation of the pfa force @xmath110 , a value of the metaphysical parameter @xmath111 m is used .
( right ) ratio @xmath99 versus the range of the yukawian force for different values of the metaphysical parameter @xmath26 and a radius of curvature of @xmath46m.,title="fig:",scaledwidth=45.0% ] for the comparison of the epfa and pfa schemes for the multilayered and corresponding homogeneous situation ( obtainable by using eqs .
( 14 ) and ( 15 ) with @xmath96 , @xmath97 , and replacing @xmath2 with @xmath98 ) .
( left ) ratio @xmath99 versus the range of the yukawian force for different values of the radius of the inner sphere .
the parameters for the layered sphere are @xmath100 nm , @xmath101 nm , @xmath28=4.1 g/@xmath102 , @xmath103=7.14 g/@xmath102 , and @xmath104=19.28 g/@xmath102 . the parameters for the layered slab are @xmath105 m , @xmath106 nm , @xmath107 nm , @xmath27=2.33 g/@xmath102 , @xmath108=7.14 g/@xmath102 , and @xmath109=19.28 g/@xmath102 . in the evaluation of the pfa force @xmath110 , a value of the metaphysical parameter @xmath111 m is used .
( right ) ratio @xmath99 versus the range of the yukawian force for different values of the metaphysical parameter @xmath26 and a radius of curvature of @xmath46m.,title="fig:",scaledwidth=45.0% ] on the other hand , the pfa expression for the force between the layered infinite slab and the layered sphere is @xmath112 , where @xmath113 is the pressure between two parallel layered slab , one identical to the previous slab , and a metaphysical slab of width @xmath26 and density @xmath28 , covered by two layers of widths and densities identical to the ones of the layered sphere above . using eq .
( [ pressure ] ) for the various pairs of layers in the different slabs , we calculate the pfa expression for the layered sphere - slab force @xmath114 + \right .
\nonumber \\ & & ~~ \rho'_1 ( 1-e^{-\delta'_1/\lambda } ) \left [ \rho_2 e^{-(\delta''_1+\delta'_2+\delta''_2)/\lambda } ( 1-e^{-d_2/\lambda } ) + \rho'_2 e^{-(\delta''_1+\delta''_2)/\lambda } ( 1-e^{-\delta'_2/\lambda } ) + \rho''_2 e^{-\delta''_1/\lambda } ( 1-e^{-\delta''_2/\lambda } ) \right ] + \nonumber \\ & & ~ \left .
\rho''_1 ( 1-e^{-\delta''_1/\lambda } ) \left [ \rho_2 e^{-(\delta'_2+\delta''_2)/\lambda } ( 1-e^{-d_2/\lambda } ) + \rho'_2 e^{-\delta''_2/\lambda } ( 1-e^{-\delta'_2/\lambda } ) + \rho''_2 ( 1-e^{-\delta''_2/\lambda } ) \right ] \right\ } .
\label{udeltapfa } \end{aligned}\ ] ] to assess the effect of the multilayered structures , we have evaluated the ratio @xmath99 with @xmath115 and @xmath116 , as a function of @xmath22 for three radii of curvatures of the sphere , assuming for the pfa calculation a value of the metaphysical parameter @xmath117 m ( see fig .
the effect of multilayers is to slightly flatten @xmath118 as compared to @xmath43 in the homogeneous case ( fig .
the dependence of the same ratio for a fixed value of @xmath2 and different values of the metaphysical parameter @xmath26 is shown in fig .
note that @xmath118 is independent of the sphere - slab separation , just as @xmath43 is .
as discussed in the introduction , the pfa used in all recent sphere - plane casimir experiments for the casimir theory - experiment comparison is expected to approximate the exact casimir force within 0.1 @xmath119 .
this expectation comes from recent analytical approaches to the sphere - plane casimir interaction @xcite that , although formally exact , require the evaluation of the determinant of an infinite - dimensional matrix , which becomes a numerically demanding task , especially in the pfa regime , @xmath6 , where larger and larger matrices are needed for convergence .
numerical computations of the exact , zero - temperature sphere - plane casimir force using parameters for metallic spheres ( @xmath120 m and optical response modeled by the simple plasma model with plasma wavelength @xmath121 nm ) show that deviations from pfa can be as large as @xmath122 for the smallest @xmath123 studied numerically ( see fig.2 of @xcite ) .
an extrapolation to smaller values of @xmath124 using a cubic polynomial fit of the numerical data is also provided in @xcite .
assuming one can use it for the recent casimir sphere - plane experiment @xcite ( with a radius of curvature @xmath125 m ) , gives a deviation from pfa of the order of @xmath7 at the smallest value of @xmath126 reached in the experiment ( @xmath127 nm ) . since the limits to non - newtonian forces
are obtained using the _ residuals _ in the casimir theory - experiment comparison , in order to meaningfully replace the exact formula of the yukawa force with its pfa approximation , the level of accuracy between these two should be therefore a small fraction , for instance 10 @xmath119 , of the accuracy with which the casimir force is controlled by using pfa rather than the exact expression for the sphere - plane casimir force .
if this condition is not fulfilled , the derived limits could be off also by a large , order of 100@xmath119 , correction .
however , targeting a @xmath128 accuracy level with respect to the casimir theory - experiment accuracy implies deviations from @xmath129 of @xmath130 , which can be obtained , as seen in fig .
2 , only in the range of @xmath22 below 100 nm .
the presence of substrates with different densities tends to mitigate the discrepancy between the epfa and the pfa , as seen by the curves in fig . 3 , but there is an irreducible systematic factor even at small @xmath22 .
indeed , in the limit @xmath131 , we have @xmath132 , that , in the case of the experiment reported in @xcite , is equal to 1.00126 , _
i.e. _ a correction already equal to 0.126 @xmath119 .
all these systematic sources of uncertainty could be even larger in experiments for which the radius of curvature of the sphere is not adequately optimized .
indeed , the use of spheres with smaller radius of curvature is affected more by this effect , as emphasized in the left plot of fig . 2 and in fig . 3 for the cases of @xmath133 m and @xmath134 m
moreover , for small spheres the pfa approximation to the casimir force itself is less accurate .
the use of spheres with large radius of curvature is beneficial to reduce these sources of error in the experiment - theory comparison , but may face experimental issues recently identified in @xcite and interpreted as due to deviations from an ideal spherical geometry ( as proposed in @xcite ) and/or a consequence of larger sensitivity to electrostatic patch effects @xcite .
in the previous section we have seen that translational invariance is crucial to make the epfa reproduce the exact result , but in actual experiments such an invariance is obviously satisfied only approximately , leading to an additional source of systematic error related to the finite size of the surfaces , as we discuss here for both power - law forces , and for yukawa forces . instead of computing the more involved problem of the gravitational force between a sphere and a finite - size slab ,
we consider here the simpler case of the gravitational force acting on a point - like test mass @xmath16 above the center of a disk of thickness @xmath25 , radius @xmath135 , and mass density @xmath27 .
we obtain @xmath136^{3/2 } } \nonumber
\\ & = & -2\pi g \rho_1 m_2 \ { d_1 + { ( r_d^2+z_2 ^ 2)}^{1/2}-{[r_d^2+(z_2+d_1)^2]}^{1/2 } \ } , \end{aligned}\ ] ] where @xmath137 is the distance between the test mass and the disk . the force becomes independent of @xmath135 only in the limit @xmath138 ( in which case it is also independent of @xmath137 ) . in order to assess the different forces acting on the various parts of a sphere in the presence of a disk of finite radius we evaluate the ratio between the forces exerted at the point of the sphere closest to the plane ( @xmath139 ) and the farthest point ( @xmath140 )
this quantity is simple to evaluate yet provides a practical figure of merit for how much the extended geometry of the sphere is affected by the finite size of the disk .
this gives a ratio @xmath141 : @xmath142^{1/2}-[(\gamma+\beta)^2+\kappa^2]^{1/2 } } { \beta+[(2+\gamma)^2+\kappa^2]^{1/2}-[(2+\gamma+\beta)^2+\kappa^2]^{1/2}},\ ] ] where we have defined @xmath143 , @xmath144 , and @xmath145 . between the force @xmath146 exerted at the closest point of the sphere from the disk and the force exerted at the farthest point , versus the radius of the disk @xmath135 constituting the planar surface of finite size , for four different exponents @xmath147 .
we assume a radius of the sphere equal to @xmath46 m , a sphere - plane distance of @xmath3=100 nm , and that the center of the sphere is right above the center of the disk .
only the case of @xmath148 ( newtonian gravitation ) gives a ratio of unity in the large @xmath135 limit .
( right ) plot of the ratio @xmath149 versus the exponent of the power force law @xmath147 for various radii of the disk @xmath135 .
the cross indicates the ratio @xmath150 for the case of an exponent @xmath148 and an infinite plane ; it is provided as a help to the eye to better show the convergence in the case of @xmath148 of @xmath149 to unity with disks of progressively larger radii.,title="fig:",scaledwidth=45.0% ] between the force @xmath146 exerted at the closest point of the sphere from the disk and the force exerted at the farthest point , versus the radius of the disk @xmath135 constituting the planar surface of finite size , for four different exponents @xmath147 .
we assume a radius of the sphere equal to @xmath46 m , a sphere - plane distance of @xmath3=100 nm , and that the center of the sphere is right above the center of the disk .
only the case of @xmath148 ( newtonian gravitation ) gives a ratio of unity in the large @xmath135 limit .
( right ) plot of the ratio @xmath149 versus the exponent of the power force law @xmath147 for various radii of the disk @xmath135 .
the cross indicates the ratio @xmath150 for the case of an exponent @xmath148 and an infinite plane ; it is provided as a help to the eye to better show the convergence in the case of @xmath148 of @xmath149 to unity with disks of progressively larger radii.,title="fig:",scaledwidth=45.0% ] this is a large correction , of the order of 300@xmath119 , if a disk of radius equal to twice the radius of the sphere ( @xmath151 ) , in a geometrical setting not dissimilar from the one used in @xcite , is considered .
since the experiment in @xcite is anyway insensitive to the gravitational force , like any experiment performed in the micrometer range , this is not a major practical concern .
however , in the case of a more generic power law such as @xmath152 we get : @xmath153^{(3-n)/2 } \ } , \ ] ] apart from the cases of @xmath154 and @xmath155 in which logarithmic integrations occur . in these two cases one obtains @xmath156 \ln[(z+d_1)^2+r_d^2]+(z+d_1)^2 \ln ( z+d_1)^2-z^2 \ln z^2 \ } , \ ] ] which in the limit @xmath157 becomes independent of @xmath34 , and @xmath158}\right ] , \ ] ] which in the limit @xmath157 behaves as @xmath159 .
notice that the fact that the force is independent on the distance from an infinite plane is only characteristic of forces scaling with the inverse square of the distance , such as the gravitational force , making the integration of the force trivially geometrical .
in the general case @xmath160 , even in the situation of a sphere in front of an infinite plane , different points of the sphere will feel different forces , with the farthest point feeling smaller ( larger ) force for a power law exponent larger ( smaller ) than 2 , as a consequence of the interplay between the solid angle and the distance scaling of the force , which makes peculiar the @xmath148 case as expressed by the gauss law .
this is shown in fig . 4 ( left ) for the cases of @xmath161 and @xmath162 , with the ratio @xmath149 between the forces evaluated at the top and at the bottom of the sphere , and in fig .
4 ( right ) by showing the same ratio versus the power law exponent for different values of the radius of the disk . therefore , when considering power - law forces as the ones discussed for instance around eq . 2 in @xcite , one should then take carefully into account the finite size of the plane in deriving limits to these forces @xcite . between the yukawa force exerted at the closest point of the sphere from the disk and the force exerted at the farthest point , versus the radius of the disk @xmath135 constituting the planar surface of finite size , for three different values of the yukawa range @xmath22 .
as before , we assume a radius of the sphere equal to @xmath46 m , a sphere - plane distance of @xmath3=100 nm , and the center of the sphere right above the center of the disk .
the values of @xmath22 are chosen to be 100 @xmath50 m , 500 @xmath50 m , and 1000 @xmath50 m .
in the last case the force may be considered as a long range one and the farthest point on the sphere is also contributing almost as the closest one . for @xmath22
progressively smaller than the radius of the sphere the ratio @xmath163 gets larger and larger and the dependence on @xmath135 is not appreciable.,scaledwidth=50.0% ] finally , we discuss the effect of the finite size of the planar surface in the case of yukawa forces .
the potential energy of a pointlike particle of mass @xmath16 located at height @xmath34 along the axis of a planar disk surface of radius @xmath135 , density @xmath27 , and thickness @xmath25 , is @xmath164^{1/2 } e^{-\sqrt{(z_2-z_1)^2+r^2}/\lambda } \nonumber \\ & = & -2 \pi \alpha g \rho_1 m_2 \lambda^2 \left[e^{-z/\lambda}(1-e^{-d_1/\lambda})-e^{-\sqrt{z^2+r_d^2}/\lambda}+e^{-\sqrt{(z+d_1)^2+r_d^2}/\lambda}\right ] , \end{aligned}\ ] ] and the related force is @xmath165^{1/2}r_d/\lambda+z/\lambda}}{[1+(z / r_d)^2]^{1/2}(1-e^{-d_1/\lambda } ) } + \frac{z+d_1}{r_d } \frac{e^{-[1+((z+d_1)/r_d)^2]^{1/2}r_d/\lambda+z/\lambda}}{[1+((z+d_1)/r_d)^2]^{1/2}(1-e^{-d_1/\lambda } ) } \right\ } , \end{aligned}\ ] ] where the finite size terms appear as corrections to the indefinite plane formula originating by the first term alone . as before , we introduce as figure of merit the ratio @xmath166 . this ratio is very large for realistic configurations , expressing the short - range nature of the force . indeed , even in the infinite plane limit we have a ratio of @xmath167 in the case of a sphere of radius @xmath46 m at a @xmath168 m .
the dependence on the disk radius become significant only at values of @xmath22 comparable to the radius of the sphere , as shown in fig .
the presence of suppression factors for the farthest point of the form @xmath169 makes very insensitive the yukawa force to the finite size of the disk ( for previous considerations , see also @xcite ) .
our analysis , although confirming some of the results already discussed in @xcite , draws quite different conclusions from the common outcome .
in particular , we argue that the application of pfa to volumetric forces is not rigorous if considered in its original formulation applied so far to compute the sphere - plane yukawa interactions in the most sphere - plane casimir force experiments reported in @xcite .
it application to volumetric forces is instead of trivial nature if considered in the exact formulation efpa discussed in @xcite , since the latter is identical to the exact calculation , just differing in the choice of the infinitesimal integration volume .
we have shown that the usual pfa is an invalid approximation to compute volumetric forces .
in particular , it does not reproduce in its usual range of validity exact known expressions for gravitational and yukawa interactions in non - translational invariant geometries , such as sphere - finite size slab or sphere - sphere configurations .
exact " pfa is the exact expression for any additive two - body interaction when one of the bodies is translational invariant , as is the case for a sphere in front of an infinite homogeneous slab or half - space . for non - translational invariant geometries , epfa also fails to give the exact result for volumetric interactions , even in the regime of parameters where it is assumed to be valid , therefore also being an invalid approximation for volumetric forces .
the difference between the two formulations of the pfa is shown to affect significantly the limits obtained so far unless one considers a regime of yukawa range so small , @xmath32 , that the approximation of a surface force ( _ i.e. _ neglecting the yukawa force due to the atoms in the `` bulk '' of the two bodies , therefore manifestly of non - volumetric character ) holds . by using the pfa the yukawa force is overestimated and therefore the limits in the @xmath0 plane become more stringent than by using the exact force estimated via epfa .
moreover , the use of pfa instead of the epfa for the parameters of the experiment supposed to provide the strongest limits to yukawian interactions @xcite occurs with an accuracy of the same order of magnitude with which the exact casimir force is expected to be also approximated by the corresponding pfa . on one hand
, since the casimir theory - experiment comparison provides force _ residuals _ that are in turn compared against the theory of yukawa forces to obtain limits on them , the use of these _ subsequent _ pfa approximations of comparable level of approximation provides a possible source of systematic error , not carefully accounted for so far . on the other hand , both casimir and yukawa pfas overestimate the respective exact forces , and therefore the systematic source of error introduced by their use might be less critical than expected on first principles . in any case , it is important to assess as much as possible both sources of errors or , in alternative , to use exact expressions for the yukawa and casimir forces . furthermore , a systematic source of error present even in the epfa scheme for finite - size planar surfaces has been discussed and shown to be significant only for power - law forces .
our analysis suggests that future limits ( or reanalysis of experiments already performed ) on yukawian forces should rely upon the use of the exact expression for the yukawa force , as performed in @xcite .
it is also worth to point out that all the above considerations hold provided that the simple scenario of additive forces is assumed , which is valid in general for weak forces among atoms in the low density limit , such that correlations leading to fluctuating forces are negligible .
however , the hypothetical yukawa forces should be located in a regime of coupling constants intermediate between the gravitational ( additive ) force and the casimir ( non additive ) force .
it is not understood a priori if the yukawa force is weak enough to make the additivity assumption reliable , and this should be kept in mind in future broad - range searches of these forces . finally ,
considering the complications emerging in the sphere - plane geometry due to the presence of previously unidentified systematics such as the sensitivity to deviations from the ideal spherical geometry @xcite and possible effects of variability of the contact potential with distance @xcite , it may be worth to focus future casimir experiments to set bounds on extra - gravitational forces on the actual parallel - plane geometry , without the drawbacks of a virtual mapping from the sphere - plane geometry made explicit in this paper .
the stronger force signal expected for the same distance between the two surfaces , the reduced sensitivity to distance - dependent contact potentials due to image charges , the absence of deviations from a uniform radius of curvature , the existence of exact mode summation techniques to compute the casimir force , the possibility to control parallelism using recently developed technology @xcite , and the possibility to compensate off - line the lack of parallelism by using the pfa as discussed in @xcite , all point in the direction to continue this class of experiments in the actual parallel plane configuration , extending below the 10 @xmath50 m range the results of the experiments described in @xcite . with regard to power - law forces
, we would like also to remark that , in the casimir theory - experiment comparison at the basis of the limits to non - newtonian forces , it is also important to calculate the exact casimir force in the presence of a planar slab of _ finite _ size , especially for the configuration used in @xcite . | we discuss the role of the proximity force approximation in deriving limits to the existence of yukawian forces - predicted in the submillimeter range by many unification models - from casimir force experiments using the sphere - plane geometry .
two forms of this approximation are discussed , the first used in most analysis of the residuals from the casimir force experiments performed so far , the second recently discussed in this context in r. decca _
et al . _ , [ phys .
rev .
d * 79 * , 124021 ( 2009 ) ] .
we show that that the former form of the proximity force approximation overestimates the expected yukawa force and that the relative deviation from the exact yukawa force is of the same order of magnitude , in the realistic experimental settings , as the relative deviation expected between the exact casimir force and the casimir force evaluated in the proximity force approximation .
this implies both a systematic shift making the actual limits to the yukawa force weaker than claimed so far , and a degree of uncertainty in the @xmath0 plane related to the handling of the various approximations used in the theory for both the casimir and the yukawa forces .
we further argue that the recently discussed form for the proximity force approximation is equivalent , for a geometry made of a generic object interacting with an infinite planar slab , to the usual exact integration of any additive two - body interaction , without any need to invoke approximation schemes .
if the planar slab is of finite size , an additional source of systematic error arises due to the breaking of the planar translational invariance of the system , and we finally discuss to what extent this may affect limits obtained on power - law and yukawa forces . |
in our universe , dark matter halos evolve in the hierarchical fashion . smallest microhalos form first , and they merge with each other to form larger halos .
the size of the smallest microhalos is determined by the scale of collisional damping and free streaming of dark matter particles .
analytical studies @xcite predicted their mass to be @xmath3 , though uncertainty in theory of supersymmetry enlarges the interval further .
early studies @xcite suggested that a significant fraction of microhalos born in early universe have survived up to present time , and they might be observed as the dominant sources of the annihilation signal .
these microhalos could enhance the annihilation signal by a factor of 2 to 5 @xcite , whose signature might have already been observed as electron and positron excess [ pamela @xcite , atic @xcite , ppb - bets @xcite , and fermi @xcite ] .
@xcite simulated the formation of microhalos using @xmath4-body simulations .
they argued that the density profiles were well fitted by a single power law , @xmath5 , with slope @xmath6 in the range of 1.5 to 2 down to the radius of @xmath7 pc and that most of microhalos will survive against galactic tidal field and encounters with stars . on the other hand
, @xcite argued that the fraction of the mass in subhalos in the solar neighborhood was significantly lower than that averaged over the entire halo .
they assumed that the spatial distribution of the microhalos is the same as that of subhalos with mass @xmath8 , and concluded that the microhalos have a negligible impact on detectability .
however , it is not clear whether their result for subhalos with a mass @xmath8 can be used to estimate the distribution of halos of @xmath9 .
since earth - mass subhalos contain no substructures by definition , their central structures can be completely different from that of more massive halos which contain many substructures . however , there is no simulation of earth - mass halos with sufficient resolution to study the their central structure , so far .
@xcite used mass resolution of @xmath10 , which was too low to determine the central structures of microhalos . in this letter
, we report the result of cosmological @xmath4-body simulations with 100 times better mass resolution and 20 times better spatial resolution compared to that used in the previous work @xcite .
we performed two high resolution cosmological @xmath4-body simulations .
the number of particles is @xmath11 .
the size of the simulation box is 30 comoving parsecs with periodic boundary condition .
the mass of particles is @xmath12 . for the time integration
, we used the greem code @xcite .
we used a leapfrog integrator with shared and adaptive time steps .
the step size was determined as @xmath13 ( minimum of these two values for all particles ) .
we simulated them from @xmath14 to @xmath15 .
the ( plummer ) softening length @xmath16 was constant in the comoving coordinate from @xmath14 to @xmath17 , and constant in the physical coordinate ( @xmath18 pc ) from @xmath17 to @xmath19 . to generate the initial particle distribution
, we used the mpgrafic package @xcite .
we considered two models with different initial matter power spectra . the matter power spectrum in model a includes the sharp cutoff corresponding to dark matter particle with a mass of 100gev @xcite .
the power spectrum of model b is without cutoff .
the cosmological parameters adopted are based on the concordance cosmological model [ wmap1 @xcite , @xmath20 . however , we used @xmath21 to be close to recent observational ones @xcite .
we did not put any thermal velocities in initial setup of model a to avoid unphysical density fluctuations in small scales @xcite .
figure [ fig : snapshot ] shows the snapshots of our simulated universe at @xmath19 .
we can see that there are no substructures in one halo of model a , except for caustics generated by non - linear growth of long - wavelength fluctuations .
this structure is quite similar to what we see in warm dark matter simulation ( e.g. * ? ? ?
* ; * ? ? ?
* ) , and different from that of galaxy - sized or larger dark - matter halos @xcite .
the difference comes simply from the initial condition .
our microhalos do not contain smaller fluctuations inside . for our standard model ( a ) and model without free - streaming cutoff ( b ) .
the width of the images correspond to 12 comoving pc .
images in the squares are enlargements of single halos .
the size of the squares is 0.6 comoving pc ( 4900aus in physical units ) .
, width=491 ] .
four panels show the density , velocity dispersion , phase - space density , and accumulated mass .
circles show profiles of a low resolution simulation ( a512 : @xmath22 particles , @xmath23 pc softening length ) .
two out of three profiles ( middle and bottom ) are vertically shifted downward by 1 and 2 dex . in the panel of phase - space density profiles ,
the two profiles ( middle and bottom ) are shifted by 2 and 4 dex .
, title="fig:",width=302 ] .
four panels show the density , velocity dispersion , phase - space density , and accumulated mass .
circles show profiles of a low resolution simulation ( a512 : @xmath22 particles , @xmath23 pc softening length ) .
two out of three profiles ( middle and bottom ) are vertically shifted downward by 1 and 2 dex . in the panel of phase - space density profiles ,
the two profiles ( middle and bottom ) are shifted by 2 and 4 dex .
, title="fig:",width=302 ] .
four panels show the density , velocity dispersion , phase - space density , and accumulated mass .
circles show profiles of a low resolution simulation ( a512 : @xmath22 particles , @xmath23 pc softening length ) .
two out of three profiles ( middle and bottom ) are vertically shifted downward by 1 and 2 dex . in the panel of phase - space density profiles ,
the two profiles ( middle and bottom ) are shifted by 2 and 4 dex .
, title="fig:",width=302 ] .
four panels show the density , velocity dispersion , phase - space density , and accumulated mass .
circles show profiles of a low resolution simulation ( a512 : @xmath22 particles , @xmath23 pc softening length ) .
two out of three profiles ( middle and bottom ) are vertically shifted downward by 1 and 2 dex . in the panel of phase - space density profiles ,
the two profiles ( middle and bottom ) are shifted by 2 and 4 dex .
, title="fig:",width=302 ]
figure [ fig : profile ] shows the spherically - averaged structures of three most massive halos .
we also performed a low resolution simulation of model a for convergence check ( the number of particles was @xmath22 and the softening length was @xmath24 pc ) .
two results are identical outside the softening radius .
the central density slope for model a is considerably steeper than that for model b. the result of model b is similar to that of large - scale cosmological simulations @xcite .
the result of model a is consistent with being a single power law of @xmath1 , for quite a wide range in the radius .
our result is reliable down to @xmath25 pc , where the density reaches more than @xmath26 .
the left - lower panel of figure [ fig : profile ] shows the rough estimate of the phase - space density given by @xmath27 , where @xmath28 is the velocity dispersion .
the slope is close to @xmath29 for @xmath30 pc .
because of liouville s theorem , this phase space density can not exceed the initial value at the time of decoupling , which is @xmath31 .
thus , the gravitational collapse of the cusp stops at the radius at which the phase space density reaches this limit .
this radius and density there are @xmath32 and @xmath33 .
we , therefore , can safely conclude the density profile of microhalos is @xmath34 and @xmath35 for @xmath36 .
this profile is consistent with the results of early studies of the collapse of self - gravitating gaseous spheres @xcite and recent high resolution simulations of cold collapse @xcite and warm dark matter @xcite .
the average density of these earth - mass halos is affected by density fluctuations of larger scales .
however , our simulations do not contain large scale fluctuations . as seen in figure [ fig : psfunc ]
, press - schechter @xcite theory predicts that the average formation epoch of microhalos is earlier if we take into account larger scale fluctuations .
since the average density of a microhalo should be proportional to the cube of its formation redshift , most microhalos should have higher average density than these in our calculation . for simplicity
, we use the profile of equation ( [ eq : profile ] ) to estimate the chance of survival of these halos .
the tidal cutoff radius of microhalo with the density profile of equation ( [ eq : profile ] ) due to galactic tide is expressed as @xmath37 , where @xmath38 is the distance from the galactic center , if we assume an isothermal halo with the rotation velocity of 220 km / s .
we can see that these halos can easily survive at the distance 1kpc , and even at the distance 0.1kpc , the central @xmath25 pc of the halo would still survive
. can encounters with stars destroy the internal cusps of microhalos @xcite ?
since the surface number density of disk stars is about 100 @xmath39 at the solar neighborhood , the closest distance a halo with the orbital period of the order of 100 myrs can approach to a star in the hubble time is @xmath40 pc .
figure [ fig : encounter ] shows the result of encounter with a @xmath41 star moving at @xmath42 .
we selected the most massive microhalo at @xmath19 from the simulation of model a ( the cutoff radius is two times virial value ) , and added velocities on each particle according to the impulsive approximation .
then , we simulated the evolution of the microhalo for 27myr after the encounter .
we can see that the regions of radius @xmath7 pc survives after encounters with the impact parameters of @xmath43 pcs . theoretically , the cutoff radius due to encounter for the power - law cusp with slope @xmath44 is given by @xmath45 , where @xmath46 is the impact parameter .
therefore , the complete disruption requires an encounter with @xmath47 .
such close encounters are expected only at the very central regions of the galaxy ( 20pc or less from the center ) .
. only particles within the radius of 0.037 pc from the center of the halo are shown .
panel b ( top - middle ) , c ( top - right ) , d ( bottom - left ) , and e ( bottom - middle ) show the distribution of particles at 27 myrs after the encounter with a solar - mass star with the impact parameters @xmath48pc , @xmath49pc , @xmath50pc , and @xmath51pc .
the black regions in the center are caused by the initial outer cutoff of the particle distribution and are not real .
the upper right box in each panel shows the central region of the microhalo .
the central regions are located at the center ( panel a and b ) , and the positions indicated by arrows ( panel c and d ) . in panel e , the central region is destroyed completely .
panel f ( bottom - right ) shows the density profiles after the encounter . ,
title="fig:",width=204 ] . only particles within the radius of 0.037 pc from the center of the halo
are shown .
panel b ( top - middle ) , c ( top - right ) , d ( bottom - left ) , and e ( bottom - middle ) show the distribution of particles at 27 myrs after the encounter with a solar - mass star with the impact parameters @xmath48pc , @xmath49pc , @xmath50pc , and @xmath51pc .
the black regions in the center are caused by the initial outer cutoff of the particle distribution and are not real .
the upper right box in each panel shows the central region of the microhalo .
the central regions are located at the center ( panel a and b ) , and the positions indicated by arrows ( panel c and d ) . in panel e
, the central region is destroyed completely .
panel f ( bottom - right ) shows the density profiles after the encounter .
, title="fig:",width=204 ] . only particles within the radius of 0.037 pc from the center of the halo
are shown .
panel b ( top - middle ) , c ( top - right ) , d ( bottom - left ) , and e ( bottom - middle ) show the distribution of particles at 27 myrs after the encounter with a solar - mass star with the impact parameters @xmath48pc , @xmath49pc , @xmath50pc , and @xmath51pc .
the black regions in the center are caused by the initial outer cutoff of the particle distribution and are not real .
the upper right box in each panel shows the central region of the microhalo .
the central regions are located at the center ( panel a and b ) , and the positions indicated by arrows ( panel c and d ) . in panel e , the central region is destroyed completely .
panel f ( bottom - right ) shows the density profiles after the encounter .
, title="fig:",width=204 ] . only particles within the radius of 0.037 pc from the center of the halo
are shown .
panel b ( top - middle ) , c ( top - right ) , d ( bottom - left ) , and e ( bottom - middle ) show the distribution of particles at 27 myrs after the encounter with a solar - mass star with the impact parameters @xmath48pc , @xmath49pc , @xmath50pc , and @xmath51pc .
the black regions in the center are caused by the initial outer cutoff of the particle distribution and are not real .
the upper right box in each panel shows the central region of the microhalo .
the central regions are located at the center ( panel a and b ) , and the positions indicated by arrows ( panel c and d ) . in panel e , the central region is destroyed completely .
panel f ( bottom - right ) shows the density profiles after the encounter .
, title="fig:",width=204 ] . only particles within the radius of 0.037 pc from the center of the halo
are shown .
panel b ( top - middle ) , c ( top - right ) , d ( bottom - left ) , and e ( bottom - middle ) show the distribution of particles at 27 myrs after the encounter with a solar - mass star with the impact parameters @xmath48pc , @xmath49pc , @xmath50pc , and @xmath51pc .
the black regions in the center are caused by the initial outer cutoff of the particle distribution and are not real .
the upper right box in each panel shows the central region of the microhalo .
the central regions are located at the center ( panel a and b ) , and the positions indicated by arrows ( panel c and d ) . in panel
e , the central region is destroyed completely .
panel f ( bottom - right ) shows the density profiles after the encounter .
, title="fig:",width=204 ] . only particles within the radius of 0.037 pc from the center of the halo
are shown .
b ( top - middle ) , c ( top - right ) , d ( bottom - left ) , and e ( bottom - middle ) show the distribution of particles at 27 myrs after the encounter with a solar - mass star with the impact parameters @xmath48pc , @xmath49pc , @xmath50pc , and @xmath51pc .
the black regions in the center are caused by the initial outer cutoff of the particle distribution and are not real .
the upper right box in each panel shows the central region of the microhalo .
the central regions are located at the center ( panel a and b ) , and the positions indicated by arrows ( panel c and d ) . in panel
e , the central region is destroyed completely .
panel f ( bottom - right ) shows the density profiles after the encounter .
, title="fig:",width=204 ] if we extrapolate the subhalo mass function @xmath52 @xcite to the mass of microhalos , the number of microhalos in our galaxy is @xmath53 ( we assumed the mass function @xmath54 ) .
the number density of the microhalos near the solar neighborhood is @xmath55 , if the distribution of the microhalos is the same as that of the background dark matter .
this extrapolation is justified by the fact that the microhalos survive under tides from our milky way galaxy .
note that the above estimate takes into account the effect of merging between small halos and accretion to somewhat larger halo since we extrapolated the mass function obtained by @xmath4-body simulation .
the gamma - ray flux of a microhalo is given by @xmath56 where @xmath57 is the number of emitted photon per annihilation , @xmath58 is the mass of dark matter particle , @xmath59 is the interaction cross section of dark matter , and @xmath60 is the outer cutoff radius of a microhalo at which the slope of the density becomes steeper than @xmath44 .
mass loss due to the galactic tide or encounters with stars changes @xmath60 of microhalos , but does not affect @xmath61 or @xmath62 except in the cases of extremely close encounters , as we can see in figure 3 .
therefore , they have relatively minor effect on the gamma - ray luminosity .
after 99% of the mass is lost from a microhalo , it still retains nearly 50% of the luminosity .
thus , we can assume that the tidal stripping has practically no effect on the gamma - ray luminosity of our earth - mass halos .
these results supports the predictions of early analytical studies @xcite .
here , we consider how the gamma - ray flux of the microhalo depends on its formation epoch .
the average density of a halo @xmath63 reflects the cosmic density at its formation time @xcite . since the cosmic density is proportional to @xmath64 , @xmath65 , where @xmath66 and @xmath67 is the formation redshift and the typical formation redshift of microhalos in our simulations . from the conservation of the mass
, we can derive @xmath68 .
we can rewrite @xmath69 in equation ( [ eq : gammarayflux2 ] ) as @xmath70 using equation ( [ eq : profile ] ) .
therefore , @xmath71 here , we drop the dependency of the formation redshift in the logarithmic term in equation ( [ eq : gammarayflux2 ] ) , since it is rather weak ( @xmath72 ) .
the average boost factor due to the formation epoch is determined as @xmath73 where @xmath74 is the distribution function of the formation epoch . for the ps function in figure [ fig : psfunc ] ,
its value is @xmath0 1.6 .
we apply this boost factor to estimate the gamma - ray luminosity .
there are many works on whether or not subhalos and microhalos can be observed via annihilation gamma - ray ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we made the all sky map of the gamma - ray annihilation signal in our galaxy based on the new profile of the microhalo .
the observer locates at 8.5 kpc from the center of the halo along its long axis .
we consider the emissions from microhalos only .
we assume the spatial distribution of microhalos follows the mass distribution of a galaxy sized lcdm halo selected from our previous simulation @xcite .
thus , our all - sky map naturally include all subhalos with the mass larger than numerical limit ( @xmath75 ) .
in addition , destruction due to encounters with stars are taken into account .
we assumed a exponential disk with the surface density , @xmath76 , where @xmath77 is the distance from the galactic center . from the position of an ensemble of microhalos ,
the impact parameter @xmath46 of encounters with stars are calculated .
then , we use the cutoff radius @xmath78 and reduce the gamma - ray luminosity . in order to include the luminosities of nearby microhalos
, we placed randomly 2000 microhalos less than 1pc from the observer .
the distribution of formation redshift and the gamma - ray flux of microhalos are given by figure [ fig : psfunc ] and equation ( [ eq : gammarayflux3 ] ) .
figure [ fig : gamma - ray ] shows the all sky map .
many nearby microhalos are observed as pointlike sources , since their angular size is around 1. therefore , the nearby microhalos are promising sources of gamma - ray signal of the annihilation of dark matter particles .
their distance is around @xmath79 and velocity is of the order of @xmath80 .
the proper motion is as large as 0.2 deg / year
. they might be too dim to be observed by fermi .
sommerfeld effect could boost the interaction cross section of dark matter , and enhance by orders of magnitude the gamma - ray luminosity @xcite .
these microhalos might be good targets of next generation cherenkov telescope , such as cta .
bright microhalos enhance luminosities of all subhalos ( including known milky way satellite ) since subhalos contain a number of microhalos .
the brightest gamma - ray source is the integrated emission of microhalos in the galactic center , but it is much more extended compared to the result of previous studies @xcite
. these sources may be also good targets for observation .
however , baryons would have significant effect on the distribution of dark matter at the center of the galaxy .
if baryons disrupt the central cusp of the main halo @xcite , gamma - ray flux from the galactic center might be much smaller .
observational data suggests that dwarf galaxies have constant density cores ( e.g. * ? ? ?
pulsar timing measurements might be used to detect microhalos . since many of known millisecond pulsars ( msps ) are in the direction of the galactic center , the number density of microhalos around msps is higher than that of solar neighbourhood . here
, we estimate the residual of the time of arrival of a msp due to the gravitational attraction from its nearest neighbor , following @xcite .
the residual due to the constant acceleration @xmath81 in the period of @xmath82 is given by @xmath83 where @xmath84 is the speed of light , and @xmath81 is given by @xmath85 where @xmath86 , @xmath87 and @xmath38 are the gravitational constant , the mass of the microhalo , and the distance between the microhalo and the msp .
so we have @xmath88 if we assume that a microhalo lost 90% of mass by encounters with stars , the residual is about @xmath89 in twenty years .
the change of the distance in ten years is around 400au . since the change in the acceleration is proportional to the third power of the distance
, one in ten msps would show the change in the acceleration ten times larger than that of a typical msp , which should be detectable with ppta @xcite after modest improvement in the timing accuracy .
note that the mass loss of two weak encounters can be larger than that of a strong encounter @xcite . for simplicity
, we used the mass loss due to a strong encounter as estimated in subsection [ sec : encounter ] .
however , the mass loss of microhalos in the opossite direction to the galactic center is less than that near sun .
msps in the direction should show larger timing residual .
we found that earth - mass microhalos have steep central density cusps .
their central regions is not disrupted by tidal forces from the parent galaxies or stars and they survive to the present time .
our result is different from the recent claims that small - scale structure have a negligible impact on dark matter detectability @xcite .
they considerably underestimate annihilation signals because they assumed the density profile for the smallest microhalos was shallower than that we found with high resolution simulations .
we are grateful to veniamin berezinsky for helpful discussions .
numerical computations were carried out on cray xt4 at center for computational astrophysics , cfca , of national astronomical observatory of japan .
is financially supported by research fellowship of the japan society for the promotion of science ( jsps ) for young scientists . this research
is partially supported by the special coordination fund for promoting science and technology ( grape - dr project ) , ministry of education , culture , sports , science and technology , japan . | earth - mass dark matter microhalos with size of @xmath0 100 aus are the first structures formed in the universe , if the dark matter of the universe are made of neutralino . here , we report the results of ultra - high - resolution simulations of the formation and evolution of these microhalos .
we found that microhalos have the central density cusps of the form @xmath1 , much steeper than the cusps of larger dark halos .
the central regions of these microhalos survive the encounters with stars except in very inner region of the galaxy down to the radius of a few hundreds pcs from the galactic center . the annihilation signals from nearest microhalos are observed as gamma - ray point - sources ( radius less than 1 ) , with unusually large proper motions of @xmath2 degree per year .
their surface brightnesses are @xmath010% of that of the galactic center . their s / n ratios might be better if they are far from the galactic plane .
luminosities of subhalos are determined only by their mass , and they are more than one order of magnitude luminous than the estimation by @xcite : a boost factor can be larger than 1000 .
perturbations to the millisecond pulsars by gravitational attractions of nearby earth - mass microhalos can be detected by the observations of parkes pulsar timing array ( ppta ) . |
the hilbert space , @xmath8 associated with a quantum system composed of three qubits is spanned by basis vectors @xmath9 where @xmath10 or @xmath11 and @xmath12 . here
@xmath13 is the dimension of hilbert space associated with @xmath14 qubit . to simplify the notation we denote the vector @xmath15 by @xmath16and write a general three qubit pure state as @xmath17to measure the overall entanglement of a subsystem @xmath18 we shall use twice the negativity as defined by vidal and werner @xcite , and call it global negativity @xmath19 .
it is an entanglement measure based on peres - hororedecki @xcite npt ( negative partial transpose ) sufficient criterion for classifying bipartite entanglement .
the global partial transpose of @xmath20 ( eq.([1 ] ) ) with respect to sub - system @xmath21 is defined as@xmath22the partial transpose @xmath23 of an entangled state is not positive . global negativity , defined as@xmath24is an entanglement monotone and measures the entanglement of subsystem @xmath21 with its complement in a bipartite split of the composite system .
global negativity vanishes on ppt - states and is equal to the entropy of entanglement on maximally entangled states .
the @xmath0way partial transpose of @xmath25 with respect to subsystem @xmath21 is obtained by transposing the indices of subsystem @xmath21 in those matrix elements , @xmath26 , that satisfy the condition @xmath27 , where @xmath28 to @xmath2 .
a typical matrix element of the three qubit state operator @xmath20 involves a change of state of @xmath4 subsystems , where @xmath28 to @xmath2 .
for example , a matrix element involving a change of state of two subsystems looks like @xmath29 ( @xmath30 ) .
the set of two distinguishable subsystems that change state while one of the sub - systems does not , can be chosen in three distinct ways .
in general , the number of spins that are flipped to get a vector @xmath31 from the vector @xmath32 is @xmath33 .
the operator @xmath20 can be split up into parts labelled by @xmath4 ( @xmath34 ) and written as @xmath35with@xmath36here @xmath37 .
the @xmath5way and @xmath6way partial transpose with respect to qubit @xmath21 are defined as@xmath38and@xmath39
the @xmath0way negativity calculated from @xmath0way partial transpose of matrix @xmath40 with respect to subsystem @xmath21 , is defined as @xmath41where @xmath42 is the trace norm of @xmath43 . using the definition of trace norm and the fact that @xmath44 , we get@xmath45 , @xmath46 being the negative eigenvalues of matrix @xmath47 . the negativity
, @xmath48 ( @xmath49 ) , depends on @xmath0 way coherences and is a measure of all possible types of entanglement attributed to @xmath0 way coherences .
intuitively , for a system to have pure @xmath3partite entanglement , it is necessary that @xmath3way coherences are non - zero . on the other hand ,
@xmath3partite entanglement can be generated by @xmath50 way coherences , as well .
for a three qubit system , maximally entangled tripartite ghz state is an example of pure tripartite entanglement involving @xmath6way coherences .
the global negativity @xmath51 , for maximally entangled three qubit ghz state .
maximally entangled w - state is a manifestation of tripartite entanglement due to @xmath5way coherences . for pure states as well as those mixed states for which the density matrix is positive , entanglement of a subsystem is completely determined by global negativity @xmath19 and the hierarchy of negativities @xmath52 (
@xmath53 calculated from @xmath47 associated with the @xmath54 sub - system . for three qubit system , @xmath55 , @xmath56 , and @xmath19 ( @xmath49 )
quantify the coherences present in the composite system .
a natural question is , how much of global negativity comes from @xmath5way transpose and how much has its origin in @xmath6way transpose , for a given qubit ?
the operator @xmath57 in its eigen basis is written as @xmath58where @xmath59and @xmath60 ( @xmath61and @xmath62 ) are the positive ( negative ) eigenvalues and eigenvectors , respectively .
as such the negativity of @xmath23 is given by @xmath63it is easily verified that @xmath64substituting eq .
( [ 13 ] ) in eq .
( [ 12 ] ) and recalling that @xmath65 is a positive operator with trace one , we get@xmath66where @xmath67 is the contribution of @xmath0way partial transpose to @xmath19 .
the set of states that can be transformed into each other by local unitary operations lie on the same orbit and have the same entanglement as the canonical state expressed in terms of the minimum number of independent vectors @xcite .
construction of pure three qubit canonical state has been given by acin et al @xcite .
it is easily verified that for the states reducible to the canonical state by local unitary operations , although @xmath19 is invariant under local operations , @xmath68 varies under local unitary operations .
for a canonical state , @xmath69 @xmath70 , is found to lie at a minimum with respect to local unitary rotations applied to any of the three qubits .
we conjecture that for a canonical state , @xmath71 is a measure of genuine @xmath6way entanglement of three qubit system .
bipartite entanglement of qubit one with qubit two equals the negativity @xmath72 of @xmath73 where the reduced operator , @xmath74 in case no w - like tripartite entanglement is present , the bipartite entanglement of a qubit is given by @xmath75 .
for a canonical state , the measure @xmath76 contains information about the w - like as well as pairwise entanglement of qubit @xmath77
three qubit greenberger - horne - zeilinger state @xmath78is a maximally entangled state having genuine tripartite entanglement . for this state @xmath79 , and @xmath80 , for @xmath49 . on the other hand
there exists a class of tripartite states akin to maximally entangled w - state given by @xmath81for the pure state @xmath82 , @xmath83 , for @xmath49 .
bipartite entanglement of qubit one and two in the state @xmath82 is measured by the negativity of partially transposed reduced density operator @xmath84 ( @xmath85 , which is found to be @xmath86 @xmath87 the total pairwise entanglement of a qubit in w - state is twice the value of @xmath88 ( = @xmath89 and is less than @xmath76(@xmath90 ) .
the residue accounts for the w - type tripartite entanglement of the system and generates @xmath91 .
the three qubit state @xmath92has no genuine tripartite entanglement and no @xmath6way coherences as such @xmath93 .
analogous to the case of state @xmath82 , we have @xmath94 and @xmath95 . fig . 1 displays @xmath96 @xmath97 and @xmath98 as a function of parameter @xmath99 .
to decipher the interplay of genuine tripartite entanglement and entanglement generated by @xmath5way coherences , we examine the coherences of single parameter pure states @xmath100and @xmath101for a given value of @xmath99 , the state may have bipartite , genuine tripartite as well as w - type entanglement as seen by @xmath102 and @xmath103 displayed in figs . ( 2 ) and ( 3 ) .
bipartite entanglement ( @xmath94 ) of a qubit in a state @xmath104 is obtained by calculating the negativity @xmath105 of partially transposed @xmath106 defined as @xmath107 .
we may remark that the states @xmath108 and @xmath109 share the same @xmath110 with @xmath105 going to zero at @xmath111
( 2 ) and ( 3 ) also display total bipartite entanglement of a qubit , @xmath112 , total tripartite entanglement @xmath113 and w - state like entanglement @xmath114 , of a qubit in the states @xmath115 and @xmath116 , respectively .
tripartite entanglement @xmath117 , resulting from the interference of @xmath118 and @xmath119 shows that @xmath5way coherence is partially annihilated by the @xmath6way coherence and vice versa .
wooters three tangle calculated from concurrence @xcite has been used to study tripartite entanglement of three qubit system in ref .
three tangle for states @xmath120 , calculated from @xmath121 and tripartite entanglement @xmath117 are plotted as a function of @xmath99 in fig .
the two measures show similar trend and match at @xmath122 , but in general @xmath6tangle overestimates the tripartite entanglement of the three qubit composite system in comparison with @xmath123 . in summary ,
we have extended the use of ppt criteria to the case of tripartite entanglement by splitting the partially transposed three qubit state operator into matrices that contain information either about bipartite or about genuine tripartite entanglement .
the set of @xmath5way , @xmath6way and global negativities make clear distinction between the coherences and nature of entanglement present in three qubit composite states . the three qubit canonical states may be classified using @xmath124 and @xmath125 ( @xmath126 to @xmath127 , @xmath128 ) as labels .
these entanglement measures are easy to calculate using standard diagonalization routines .
we believe that the relations between @xmath0way negativities and entanglement of a composite quantum system , a ) with sub - system dimensions higher than two , and b ) with more than three subsystems , can be figured out .
we expect our results to help advance the investigation of multipartite entanglement .
financial support from national council for scientific and technological development ( cnpq ) , brazil and state university of londrina , ( faep - uel ) , brazil is acknowledged . | in this letter we propose to quantify three qubit entanglement using global negativity along with @xmath0way negativities , where @xmath1 and @xmath2 . the principle underlying the definition of @xmath0way negativity for pure and mixed states of @xmath3subsystems is ppt sufficient condition .
however , @xmath0way partial transpose with respect to a subsystem is defined so as to shift the focus to @xmath0way coherences instead of @xmath4 subsystems of the composite system .
[ multiblock footnote omitted ] quantum entanglement is not only a fascinating aspect of multipartite quantum systems , but also a physical resource needed for quantum communication , quantum computation and information processing in general .
bipartite entanglement is well understood , however , many aspects of multipartite entanglement are still to be investigated
. peres @xcite and the horedecki @xcite have shown a positive partial transpose ( ppt ) of a bipartite density operator to be a sufficient criterion for classifying bipartite entanglement .
negativity zycz98,vida00 based on peres horodecski criterion has been shown to be an entanglement monotone @xcite .
negativity is a useful concept being related to the eigenvalues of partially transposed state operator and can be calculated easily . in this letter
, we define @xmath5way and @xmath6way negativities and propose a classification of three qubit states based on measures related to global , @xmath5way and @xmath6way negativities .
general definition of @xmath0way negativities for pure and mixed states of @xmath3subsystems is given in ref @xcite .
the @xmath0way partial transpose with respect to a subsystem is defined so as to shift the focus to @xmath0way coherences instead of @xmath4 subsystems of the composite system . while pure @xmath0 partite entanglement of a composite system is generated by @xmath0way coherences , @xmath0partite entanglement can ,
in general , be present due to @xmath7 way coherences as well . |
we use the resonance lines of h - like o and of h - like and he - like ne to estimate the ne / o abundance ratio . in hot ( @xmath10-@xmath11 k ) coronal plasma these lines
are formed predominantly by radiative de - excitation of levels excited by collisions with thermal electrons .
the flux , @xmath12 , from such a transition @xmath13 in an ion of an element with abundance @xmath14 can be written as @xmath15 \;dt % \overline{n_e^2}(t)\ , \frac{dv(t)}{dt } \;dt \,\,\ , \mbox{erg cm$^{-2}$ s$^{-1}$ } \label{e : flux}\ ] ] where @xmath16 describes the line _ emissivity_the product of the relative population of the ion in question and the excitation rate of the transition as a function temperature , @xmath17 .
the kernel @xmath18the emission measure distribution describes the excitation power of the plasma as a function of temperature , which is proportional to the mean of the square of the electron density , @xmath19 , and the emitting volume @xmath20 , @xmath21 .
if we can choose o and ne lines whose @xmath16 functions have very similar temperature dependence , an abundance ratio by number , @xmath22 , can be derived simply from the ratio of their observed line fluxes , @xmath23 and @xmath24 , since all the temperature - dependent terms in equation [ e : flux ] cancel : @xmath25 an early study of ne / o ratios in solar active regions@xcite used the ratio of ne ix @xmath26 to o viii @xmath27 .
this ratio does , however , have some significant residual dependence on temperature.@xcite here we remove much of this temperature dependence by addition of ne x @xmath28 ; our combined ne @xmath16 function is @xmath29 . the resulting ratio @xmath30 is illustrated as a function of temperature in figure [ f : emissrat ] .
we have verified the small residual temperature sensitivity evident in the lower panel of figure [ f : emissrat ] to be negligible for our analysis by integrating the products of @xmath31 and @xmath32 with empirically - derived emission measure distributions , @xmath18 , for different stars,@xcite and for functions @xmath33 , with @xmath34 : the integrated emissivity ratio from these tests was @xmath35 .
we conclude that the line ratio method is robust and the higher ne / o abundance ratio found here will not be significantly changed through performing full emission measure distribution modelling .
we measured ne and o line fluxes ( listed in table 1 ) from _ chandra _ hetg x - ray spectra obtained directly from the chandra public data archive ( http://cda.harvard.edu ) .
final listed fluxes for ne x include small reductions ( @xmath36% for 17 out of 21 or our stars , and 25 - 37% for the remainder ) to account for a weak blend of fe xvii at 12.12 .
the fe xvii 12.12 contribution was estimated by scaling the observed strengths of unblended fe xvii lines at 15.26 , 16.77 , 17.05 and 17.09 ( the strong 15.01 resonance line was omitted to avoid potential problems with its depletion through resonance scattering ) by their theoretical line strengths relative to the 12.12 line as predicted by the chianti database .
minor blending in the wings of the ne ix 13.447 line was accounted for by fitting simultaneously with the neighbouring weaker lines , comprised of a fe xix - xxi blend at 13.424 and fe xix 13.465 , following a detailed study of these features in the capella binary system.@xcite since these blend corrections are generally very small , the uncertainties in these procedures have negligible ( @xmath37% ) influence on the derived ne / o abundance ratios .
jjd was supported by a nasa contract to the _ chandra x - ray center_. pt was supported by a chandra award issued by chandra x - ray center , which is operated by sao for and on behalf of nasa .
jjd thanks the nasa aisrp for providing financial assistance for the development of the pintofale package .
we thank drs .
g. share , r. murphy , w. ball and d.garcia-alvarez for useful discussions and comments .
.spectral line fluxes and derived ne / o abundance ratios for the stars analysed in this study .
line fluxes were measured from the medium energy grating ( meg ) component of _ chandra _ hetg spectra by line profile fitting using the package for interactive analysis of line emission ( pintofale ) software@xcite ( freely available from http : hea - www.harvard.edu / pintofale ) .
the effective collecting area of the instrument was accounted for using standard _
chandra _ calibration products and techniques ( see http://cxc.harvard.edu/ciao/ for details ) .
ne / o abundance ratios were obtained assuming the o / ne line emissivity ratio of @xmath38 , as described in methods . stated flux and
abundance ratio uncertainties correspond to @xmath39 limits .
[ cols="<,^,^,^,^,^,^ , < " , ] , vs. the coronal activity index @xmath1 .
error bars represent quadrature addition of @xmath39 uncertainties of line flux measurement .
also shown using hollow symbols are literature values@xcite for the stars procyon ( f5 iv ) and @xmath2 eri ( k2 v ) observed using the _ chandra _ low energy transmission grating spectrometer ( letgs ) to better represent the lower ranges of coronal activity .
the error - weighted mean ne / o abundance ratio is @xmath41 , or 2.7 times the currently assessed value@xcite which is illustrated by the dashed horizontal line .
the recommended value from comprehensive earlier assessments in common usage@xcite are also illustrated.,scaledwidth=100.0% ] , of the o viii @xmath28 line , and @xmath32 of the ne ix @xmath42 and ne x @xmath28 lines combined as @xmath29 .
the lower panel shows the logarithmic ratio @xmath30 .
emissivities are based on electron excitation rates and ion populations@xcite compiled in the chianti database,@xcite as implemented in pintofale.@xcite , scaledwidth=80.0% ] | the interior structure of the sun can be studied with great accuracy using observations of its oscillations , similar to seismology of the earth .
precise agreement between helioseismological measurements and predictions of theoretical solar models@xcite has been a triumph of modern astrophysics . however , a recent downward revision by 25 - 35% of the solar abundances of light elements such as c , n , o and ne@xcite has broken this accordance :
models adopting the new abundances incorrectly predict the depth of the convection zone , the depth profiles of sound speed and density , and the helium abundance.@xcite the discrepancies are far beyond the uncertainties in either the data or the model predictions.@xcite here we report on neon abundances relative to oxygen measured in a sample of nearby solar - like stars from their x - ray spectra .
they are all very similar and substantially larger than the recently revised solar value .
the neon abundance in the sun is quite poorly determined .
if the ne / o abundance in these stars is adopted for the sun the models are brought back into agreement with helioseismology measurements.@xcite harvard - smithsonian center for astrophysics , 60 garden street , cambridge ma 02138 , usa .
mit kavli institute for astrophysics and space research , massachusetts institute for technology , 70 vassar street , cambridge , ma 02139 , usa .
the role of the sun as a fundamental benchmark of stellar evolution theory , which itself underpins much of astrophysics , renders the `` solar model problem '' one of some importance .
the schism between helioseismology and models with a revised composition has arisen because abundant elements such as c , n , o and ne provide major contributions to the opacity of the solar interior , which in turn influences internal structure and the depth at which the interior becomes convective .
uncertainties in the calculated opacities themselves appear insufficient to bridge the gap,@xcite and the propriety of the recent abundance revisions has therefore been questioned .
the revised abundances are demanded by new analyses of the visible solar spectrum that take into account convection and associated velocity fields and temperature inhomogeneities through 3-d hydrodynamic modelling , and that relax the assumption of local thermodynamic equilibrium for computing atomic level populations.@xcite the solar abundances of c , n and o can be measured accurately based on their absorption lines .
however , ne lacks detectable photospheric lines in cool stars like the sun , and the ne abundance is therefore much less certain .
the solar ne content is assessed based on observations of neon ions in nebular and hot star spectra , and on measurements of solar energetic particles.@xcite measurements are generally made relative to a reference element such as o ; the downward revision of the solar o abundance therefore required a commensurate lowering of the ne abundance for consistency.@xcite expressed as the ratio of the number of ne atoms in the gas to those of o , this abundance is @xmath0 , which is very similar to values adopted in earlier studies.@xcite however , it has recently been pointed out that the solar model problem could be solved were the true solar ne abundance to be at least a factor of 2.5 times higher than recently assessed.@xcite we are motivated by the solar model problem to investigate ne abundances in nearby stars . while not detected in the optical spectra of cool stars , emission lines of highly ionised ne are prominent features of their x - ray spectra.@xcite the ne / o abundance ratio can be derived directly from the ratio of observed fluxes of the hydrogen - like and helium - like ions of o and ne ( see methods for details ) ; we adopt a slightly refined version of the method applied to the analysis of earlier solar x - ray spectra.@xcite in this way , we have obtained the ne / o abundance ratios for a sample of 21 stars lying within 100pc of the sun that have been observed by the _
chandra _ x - ray observatory using the high energy transmission grating spectrometer.@xcite a representative x - ray spectrum , that of the m1 v star au mic , is presented in figure [ f : megspec ] .
the stars studied , together with observed ne and o line fluxes and derived ne / o abundance ratios are listed in table 1 .
the ne / o abundance ratios are illustrated as a function of the ratio of logarithmic x - ray and bolometric luminosities , @xmath1a commonly used index of stellar coronal activity in figure [ f : abunrats ] . since our star sample contains more objects toward higher @xmath1 , we have added ne / o ratios for two stars of somewhat lower activity level ( procyon , an f5 subgiant , and @xmath2 eri , a k2 dwarf ) from the literature.@xcite there is no trend in the ne / o abundance ratio with @xmath1 , and the error - weighted mean ratio is @xmath3 .
our derived ne / o abundance ratio is 2.7 times higher than the currently recommended solar value@xcite but is consistent with the abundance inferred from helioseismology.@xcite solving the solar model problem by raising the ne abundance alone would require a minimum ratio @xmath4 , or 3.44 times larger than recommended.@xcite however , raising the c , n , o and fe abundances upward within their estimated uncertainty range of @xmath5% ( adjusting them all together is not unreasonable because the recent downward revisions are correlated ) would require an ne abundance higher by only a factor of 2.5@xcite quite within our estimated range . extensive review articles on the coronae of the sun and stars document evidence that coronal chemical compositions often appear different to those thought to characterise the underlying photospheres.@xcite
the differences appear to relate to element first ionisation potentials ( fips ) : the solar corona appears enhanced in elements with low fip ( @xmath6 ev ; e.g. , mg , si fe),@xcite whereas much more coronally active stars appear to have depletions in low fip elements and perhaps enrichments of high fip elements ( @xmath6 ev ; e.g. , o , ne , ar).@xcite at first sight , this chemical fractionation might render direct interpretation of coronal abundance ratios in terms of the composition of the underlying star problematic .
however , that we are seeing the same ne / o ratio in a wide variety of stars sampling a large range of different coronal activity levels indicates that there is no significant fractionation between o and ne in disk - integrated light from stellar coronae .
this leads us to conclude that the results represent the true ne / o abundance ratios of these stars .
this conclusion is bolstered by findings of a constant ne / o abundance ratio , in good agreement with our value , for a small sample of single and active binary stars observed by _ xmm - newton_.@xcite in view of the consistency of the ne / o abundance ratio in nearby stars , it seems likely that the solar ratio should be similar .
there are no recent full - disk integrated light ne / o measurements for the sun ; existing studies are instead based on observations of particular regions and structures of the solar outer atmosphere . we provide a tabular summary and bibliography of some of the different ne / o estimates since 1974 as supplementary information to this letter .
while the preponderance of estimates appear consistent with current assessments , the underlying observations do not sample photospheric material .
the ne / o ratio is in fact observed to differ substantially between the different observations . in particular , the highest measured ne / o ratios based on x - ray lines are 2 - 3 times the accepted solar value and are compatible with the abundance ratio we find for nearby stars .
these higher ne / o ratios were obtained for hotter active regions@xcite that are likely to dominate the solar full - disk x - ray emission .
these measurements are the most directly compatible with the ones presented here based on full - disk integrated light x - ray spectra of stars .
similarly `` high '' ne / o ratios have also been seen in @xmath7-ray observations of flares , @xmath8he - rich solar energetic particle events , and in the decay phase of long duration soft x - ray events.@xcite the @xmath7-ray measurements probe material that is irradiated by downward - flowing accelerated particles.@xcite the particle beams penetrate through to the chromosphere which is likely more representative of the underlying photospheric material than coronal regions .
recent sophisticated models of heliospheric pickup ion and anomalous cosmic ray populations from the local interstellar medium are also inconsistent with the current solar ne / o abundance ratio , but could be reconciled by raising this to @xmath9,@xcite which is similar to our values for nearby stars .
this ne / o ratio is somewhat higher than the mean from studies of ionized nebulae in the milky way and other galaxies , but falls within the scatter of results at solar metallicity.@xcite the implication of our study , then , is that the higher of the observed solar ne / o abundance ratios are the ones representative of the underlying solar composition .
this scenario is in accordance with our observations of ne / o in nearby stars and reconciles solar models with helioseismology . |
the analytic hierarchy process ( ahp ) is a method for ranking alternatives in multi - criteria decision making problems .
developed by saaty @xcite , it consists of a three layer hierarchical structure : the overall goal is at the top ; the criteria are in the next level ; and the alternatives are in the bottom level .
the ahp has been used in many different areas including manufacturing systems , finance , politics , education , business and industry ; for more details on the method , see the monographs by saaty - vargas and vaidya - kumar @xcite .
the essence of the ahp can be described as follows . given @xmath0 alternatives we construct a _ pairwise comparison matrix _ ( _ pc_-matrix ) , @xmath1 for each criterion , in which @xmath2 indicates the strength of alternative @xmath3 relative to alternative @xmath4 for that criterion .
pc_-matrix with the property that @xmath5 for all @xmath6 and @xmath7 for all @xmath3 is called a _ symmetrically reciprocal _ matrix ( _ sr_-matrix ) @xcite .
( note that this abbreviation might clash with the strongly regular matrices of butkovi @xcite , but not in this paper . )
sr_-matrix @xmath8 is constructed , the next step in the ahp is to derive a vector @xmath9 of positive weights , which can be used to rank the alternatives , with @xmath10 quantifying the weight of alternative @xmath3 .
as observed by elsner and van den driessche @xcite , the ideal situation is where @xmath11 , in which case the _ sr_-matrix is _
transitive_. in practice , this will rarely be the case and it is necessary to approximate @xmath8 with a transitive matrix @xmath12 , where @xmath13 for some positive weight vector @xmath14 .
the problem is then how to construct @xmath12 given @xmath8 .
several approaches have been proposed including saaty s suggestion to take @xmath15 to be the perron vector of @xmath8 , or the approach of farkas et al .
@xcite , which chooses @xmath15 to minimise the euclidean error @xmath16 . elsner and van den
driessche @xcite suggested selecting @xmath15 to be the max algebraic eigenvector of @xmath8 .
this is similar in spirit to saaty s approach and also generates a transitive matrix that minimises the maximal relative error @xmath17 . as noted in @xcite , minimising this functional is equivalent to minimising @xmath18 the different approaches to approximating an _
sr_-matrix @xmath8 with a transitive matrix @xmath12 will in general produce different rankings of the alternatives .
the question of how these rankings are affected by the choice of scheme is considered in the recent paper of ngoc @xcite . in the classical ahp involving multiple criteria , a set of _
sr_-matrices is constructed : one for each criterion .
one additional _
sr_-matrix is constructed based on comparisons of the different criteria .
once weight vectors are obtained for each individual criterion , these are then combined using the entries of the weight vector for the criteria - comparison matrix . as an illustration
, we take the following numerical example from saaty @xcite and show how the perron vectors of the comparison matrices are used to construct a weight vector .
[ ex : saaty ] the problem considered is deciding where to go for a one week vacation among the alternatives : 1 .
short trips , 2 .
quebec , 3 .
denver , 4 .
five criteria are considered : 1 .
cost of the trip , 2 .
sight - seeing opportunities , 3 .
entertainment , 4 .
means of travel and 5 . dining .
pc_-matrix for the criteria and its perron vector are given by @xmath19\quad \text{and}\quad c=\left [ \begin{array}{c } 0.179 \\ 0.239 \\
0.431 \\ 0.818 \\ 0.237 \\ \end{array } \right ] .\ ] ] the above matrix @xmath20 describes the pairwise comparisons between the different _
criteria_. for instance , as @xmath21 , criterion 2 is rated more important than criterion 1 ; @xmath22 indicates that criterion 3 is rated more important than criterion 2 and so on .
the vector @xmath23 contains the weights of the criteria ; in this method , criterion 4 is given most weight , followed by criterion 3 and so on .
the _ sr_-matrices , @xmath24 , for each of the 5 criteria , their perron vectors and corresponding ranking schemes are given below .
for instance , for criterion 1 , the first alternative is preferred to the second as the @xmath25 entry of @xmath26 is @xmath27 .
similarly , for criterion 3 , the 4th alternative is preferred to the 1st as the @xmath28 entry of @xmath29 is @xmath30 .
for the cost of the trip : @xmath31 , \quad v^{(1)}=\left [ \begin{array}{c } 0.877 \\ 0.46 \\
0.123 \\ 0.064 \\ \end{array } \right ] , \quad 1>2>3>4\ ] ] for the sight - seeing opportunities : @xmath32 , \quad v^{(2)}=\left [ \begin{array}{c } 0.091 \\ 0.748 \\ 0.628 \\ 0.196 \\ \end{array } \right ] , \quad 2>3>4>1\ ] ] for the entertainment : @xmath33 , \quad v^{(3)}=\left [ \begin{array}{c } 0.57 \\ 0.096 \\ 0.096 \\ 0.81 \\ \end{array } \right ] , \quad 4>1>2=3\ ] ] for the means of travel : @xmath34 , \quad v^{(4)}=\left [ \begin{array}{c } 0.396 \\ 0.355 \\ 0.768 \\ 0.357 \\ \end{array } \right ] , \quad 3>1>4>2\ ] ] for the dining : @xmath35 , \quad v^{(5)}=\left [ \begin{array}{c } 0.723 \\ 0.642 \\
0.088 \\ 0.242 \\ \end{array } \right ] , \quad 1>2>4>3\ ] ] to obtain the overall weight vector , we compute the weighted sum @xmath36 .
this gives @xmath37\ ] ] with the associated ranking : @xmath38 .
our work here is inspired by the max - algebraic approach to the ahp introduced by elsner and van den driessche @xcite and extends it in the following manner . in @xcite , the max eigenvector is used as a weight vector for a _ single criterion _ and it is shown to be optimal in the sense of minimising the maximal relative error as discussed above .
this work naturally raises the question of how to treat multiple criteria within the max - algebraic framework .
we address this question here by considering the multi - criteria ahp as a multi - objective optimisation problem , in which we have an objective function of the form ( [ eq : errf ] ) for each criterion ( and associated _
sr_-matrix ) . rather than combining individual weight vectors as in example [ ex : saaty ] , we consider three approaches within the framework of multi - objective optimisation , and use the optimal solution as a weight vector in each case .
the advantage of this approach is that the weight vector can be interpreted in terms of the maximal relative error functions ( [ eq : errf ] ) associated with the _ sr_-matrices given as data for the problem .
the optimisation problems we consider are the following .
first , we investigate the existence of a single _ transitive _ matrix with a minimum distance to all matrices in the set simultaneously .
we remark that this amounts to finding a common subeigenvector of the given matrices .
clearly , this will not in general be possible .
the second problem we consider is to obtain a transitive matrix that minimises the maximal distance to any of the given _ sr_-matrices .
the third problem concerns the existence of a transitive matrix that is pareto optimal for the given set of matrices . to illustrate our results , we revisit example [ ex : saaty ] towards the end of the paper
the set of all nonnegative real numbers is denoted by @xmath39 ; the set of all @xmath0-tuples of nonnegative real numbers is denoted by @xmath40 and the set of all @xmath41 matrices with nonnegative real entries is denoted by @xmath42 .
we denote the set of all @xmath0-tuples of positive real numbers by @xmath43 . for @xmath44 and @xmath45
, @xmath2 refers to the @xmath46 entry of @xmath8 .
the matrix @xmath47 $ ] is nonnegative ( positive ) if @xmath48 ( @xmath49 ) for @xmath45 .
this is denoted by @xmath50 ( @xmath51 ) .
the weighted directed graph of @xmath8 is denoted by @xmath52 .
it is an ordered pair @xmath53 where @xmath54 is a finite set of nodes @xmath55 and @xmath56 is a set of directed edges , with an edge @xmath57 from @xmath3 to @xmath4 if and only if @xmath58 . a path is a sequence of distinct nodes @xmath59 of length @xmath60 with the weight @xmath61 , where @xmath62 is an edge in @xmath52 for @xmath63 .
it is standard that @xmath8 is an irreducible matrix if and only if there is a directed path between any two nodes in @xmath52 .
a cycle @xmath64 of length @xmath65 is a closed path of the form @xmath66 .
the @xmath67 root of its weight is called its cycle geometric mean . for a matrix @xmath50 , the maximal cycle
geometric mean over all possible cycles in @xmath52 is denoted by @xmath68 . a cycle with the maximum cycle geometric mean
is called a critical cycle .
nodes that lie on some critical cycle are known as critical nodes and denoted by @xmath69 .
the set of edges belonging to critical cycles are said to be critical edges and denoted by @xmath70 .
the critical matrix of @xmath8 @xcite , @xmath71 , is formed from the submatrix of @xmath8 consisting of the rows and columns corresponding to critical nodes as follows .
set @xmath72 if @xmath57 lies on a critical cycle and @xmath73 otherwise .
we use the notation @xmath74 for the critical graph where @xmath75 .
the max algebra consists of the set of nonnegative numbers together with the two basic operations @xmath76 and @xmath77 .
this is isomorphic to the max - plus algebra @xcite via the natural isomorphism @xmath78 .
these operations extend to nonnegative matrices and vectors in the obvious manner @xcite .
for @xmath8 in @xmath42 , the eigenequation in the max algebra is given by @xmath79 @xmath68 is the largest max eigenvalue of @xmath8 @xcite .
if @xmath8 is irreducible , then it is the unique max eigenvalue of @xmath8 and there is a positive max eigenvector @xmath80 corresponding to it @xcite .
the eigenvector @xmath81 is unique up to a scalar multiple if and only if @xmath71 is irreducible .
observe that an _ sr_-matrix @xmath8 is irreducible and @xmath82 @xcite .
although our primary interest is in _ sr_-matrices , it is noteworthy that many of our results also hold true for non - zero reducible matrices . for @xmath50 with @xmath83 , @xmath84 converges to a finite matrix called the kleene star of @xmath8 given by @xmath85 where @xmath86 @xcite . here , @xmath87 is the maximum weight of a path from @xmath3 to @xmath4 of any length @xcite ( if @xmath6 ) , and @xmath88 denotes the @xmath89 max - algebraic power of @xmath8 .
note that for each @xmath50 , if @xmath90 is finite then the max - algebraic sum of all of the columns of @xmath90 is positive .
for @xmath50 , the set of subeigenvectors of @xmath8 associated with @xmath68 is called a subeigencone of @xmath8 and denoted by @xmath91 @xcite .
it was shown in proposition 2.5 of @xcite that for @xmath50 with @xmath92 , @xmath93 where @xmath94 denotes the eigencone of @xmath90 consisting of its max eigenvectors .
note that the above - mentioned max - algebraic sum of all columns of @xmath90 is in @xmath95 @xcite , so @xmath95 contains positive vectors .
note that if @xmath96 , we can normalise @xmath8 by @xmath68 and @xmath97 . to the authors knowledge
, max - algebraic subeigenvectors appeared in the works of gaubert @xcite .
however , they can be traced back to earlier works on nonnegative matrix scaling , see references in butkovi - schneider @xcite .
for @xmath50 we will consider the following set , which was introduced in @xcite @xmath98 for the special case of @xmath99 , @xmath100 is denoted by @xmath101 @xcite .
obviously @xmath101 is the positive part of @xmath95 ( which is non - empty as we argued above ) , and it coincides with @xmath95 when @xmath8 is irreducible . to be consistent with the notation of @xcite ,
we recall the definition of the normalised set @xmath102 as above , @xmath103 is used to denote the special case where @xmath99 .
the relations between the sets @xmath104 , the error function ( [ eq : errf ] ) and @xmath68 were clarified by elsner and van den driessche @xcite and are recalled in the following propositions , which we easily extend ( based on @xcite ) to the general reducible case .
[ pro : carprop ] let @xmath50 be nonzero . then
: * @xmath105 @xmath106 ; * @xmath107 @xmath108 . to prove(i ) , exploit @xcite theorem 1.6.29 stating that @xmath109 when @xmath96 ( based on butkovi - schneider @xcite theorem 2.6 ) . in the trivial case @xmath110 , we have @xmath111 , which consists of all vectors @xmath112 such that @xmath113 if and only if the @xmath114 column of @xmath8 is zero . in this case @xmath115 , unless @xmath116 ( which we exclude ) .
the result of ( ii ) follows from the definitions of @xmath117 and @xmath118 given in and .
see also gaubert @xcite ch .
iv lemma 1.3.8 , krivulin @xcite lemma 1 , and nussbaum @xcite theorem 3.1 ( in a more general nonlinear context ) for closely related statements .
[ pro : unique ] let @xmath50
. then @xmath119 is unique if and only if @xmath71 is irreducible and @xmath120 .
proposition [ pro : carprop ] ( i ) implies that we have @xmath121 when @xmath122 , so we can assume @xmath96 and , further , @xmath92 . according to @xcite theorem 2.8 , @xmath123 here @xmath124 is the @xmath114 column of @xmath90 .
the subset @xmath125 is such that for each ( maximal ) irreducible submatrix of @xmath71 there is a unique index of that submatrix in @xmath126 . by the same theorem of @xcite , based on the classical results in @xcite , columns of @xmath90 with indices in the same irreducible submatrix of @xmath71
are proportional to each other , and there is no proportionality between different @xmath124 appearing in ( moreover , these @xmath124 are strongly linearly independent @xcite ) . `
if' : as @xmath71 is irreducible and all nodes are critical , all columns of @xmath90 are proportional to each other , and shows that @xmath95 is just one ray .
@xmath127 is irreducible as well , so @xmath128 and @xmath103 is the unique vector on that ray with @xmath129 .
`` only if '' : let @xmath71 be reducible , or let @xmath130 .
the vector @xmath131 is positive .
consider @xmath132 for all @xmath3 , taking small enough @xmath133 so that all entries of @xmath134 are less than any nonzero entry of @xmath90 . in this case
the positive entries of @xmath124 do not change , and if @xmath124 and @xmath135 are not proportional then neither are @xmath132 and @xmath136 .
after suitably normalising these vectors , we obtain two different vectors in @xmath103 .
next , let @xmath137 be a finite set of nonnegative matrices given by @xmath138 given @xmath139 , let @xmath140 denote the set of all products of matrices from @xmath139 of length @xmath141 .
formally , @xmath142 using this , the max version of the generalised spectral radius @xcite is defined by @xmath143 before stating the next theorem , let @xmath144 be the matrix given by @xmath145 note that @xmath146 if at least one @xmath147 .
moreover , @xmath148 if at least one @xmath149 and @xmath144 is irreducible if at least one @xmath150 is irreducible : these represent the main cases in which we are interested .
[ thm : mumspsis ] let @xmath139 be given by ( [ eq : psi ] ) and @xmath144 be given by ( [ eq : smat]).then , @xmath151 .
( gaubert @xcite , benek gursoy and mason @xcite ) inspired by the approach to the single criterion ahp adopted in @xcite , we associate a set @xmath152 with the set @xmath139 of nonnegative matrices and show how the geometric properties of @xmath104 discussed in @xcite extend to this new setting .
define @xmath153 we also consider the set of normalised vectors : @xmath154 we will use the notations @xmath155 for @xmath156 and @xmath157 for @xmath158 .
the following result shows that the set @xmath152 and the set @xmath159 are equal .
this will allow us to readily extend properties of @xmath104 established in @xcite to sets of matrices .
[ thm : subs ] consider the set @xmath160 given by ( [ eq : psi ] ) , and let @xmath144 be given by ( [ eq : smat ] ) .
then : @xmath161
( i ) : let @xmath112 in @xmath152 be given . then , @xmath162 which implies @xmath163 for each @xmath164 from ( ii ) in proposition [ pro : carprop ] . taking the maximum of both sides from @xmath165 to @xmath166 , we see that @xmath167 .
it follows that @xmath168 .
thus , @xmath169 .
now choose some @xmath170 .
then @xmath171 from ( ii ) in proposition [ pro : carprop ] .
since @xmath172 for all @xmath173 , we obtain @xmath174 .
thus , @xmath175 .
hence @xmath176 .
the following corollary extends proposition [ pro : carprop ] to a set of nonnegative matrices @xmath139 . since @xmath151 by theorem [ thm : mumspsis ] ( ii ) ,
the corollary is an immediate consequence of theorem [ thm : subs ] .
[ cor : cprprop ] consider the set @xmath160 given by ( [ eq : psi ] ) , and let @xmath144 be given by ( [ eq : smat ] ) . then : * @xmath177 , @xmath178 ; * @xmath179 @xmath180 . theorem [ thm : subs ] establishes that @xmath181 .
it is immediate that we also have @xmath182 and @xmath183 .
therefore , studying these sets for a collection of nonnegative matrices reduces to studying the sets associated with the single matrix @xmath144 .
this fact means that the properties of @xmath104 discussed in theorem 1 of @xcite can be directly extended to @xmath152 .
we state some of these in the following theorem , noting that property ( i ) does not require irreducibility .
[ thm : geop ] let @xmath184 be given in ( [ eq : psi ] ) , and let @xmath144 be given by ( [ eq : smat ] ) . then : * @xmath152 and @xmath185 are convex and max - convex ; * @xmath157 consists of only one vector if and only if @xmath186 is irreducible and @xmath187 .
* if all matrices in @xmath139 are irreducible then @xmath188 is compact .
( i ) : ( see @xcite proposition 3.1 ) @xmath152 is convex and max - convex , since @xmath189 where the sets whose intersection is taken are convex and max - convex .
@xmath185 inherits these properties as a coordinate section of @xmath152 .
\(ii ) follows from proposition [ pro : unique ] applied to @xmath144 ; ( iii ) follows from @xcite theorem 1 applied to @xmath144 .
we remark that sets that are both convex and max - convex have appeared under various names like kleene cones @xcite , polytropes @xcite , or zones @xcite .
the application of the max algebra to the ahp is motivated in @xcite by the following considerations .
first , it is observed that , for an _
sr_-matrix @xmath8 , vectors in the set @xmath101 minimise the function ( [ eq : errf ] ) and hence the relative error . based on this observation ,
these vectors are used to construct transitive matrices to obtain an overall ranking of the alternatives in the decision process . in light of the properties of @xmath101 ,
this is justified by the fact that the transitive matrices constructed in this way are _
closest _ to the original _
@xmath8 in the sense of the relative error .
thus , the approach to construct a ranking vector for a single _ sr_-matrix taken in @xcite amounts to solving the following optimisation problem .
@xmath190 in this and the following section , we are concerned with extending the above approach to the general ahp with @xmath0 alternatives and @xmath166 criteria .
formally , we are given @xmath166 _
sr_-matrices ; one for each criterion .
let @xmath139 in ( [ eq : psi ] ) denote the set of these matrices . for each @xmath164 , there is an error function @xmath191 defined as in ( [ eq : errf ] ) .
in contrast to the approach taken in the classical ahp , we view the construction of a ranking vector for the @xmath166 criteria as a multi - objective optimisation problem for the error functions @xmath192 , @xmath193 . to begin with , we seek a vector that simultaneously minimizes all of the functions @xmath192 .
such a vector is said to be a _ globally optimal solution _ for the multi - objective optimisation problem .
note that for each @xmath194 , the set of vectors that minimise @xmath195 is precisely @xmath196 @xcite : formally , @xmath197 hence , the problem of finding a vector @xmath198 that simultaneously minimises all the error functions @xmath192 amounts to determining when @xmath199 equivalently , @xmath112 simultaneously minimises all the error functions if and only if it is a common subeigenvector of @xmath150 for all @xmath200 .
the remainder of this section is divided into two parts : we first consider the existence of common subeigenvectors for arbitrary nonnegative matrices in the next subsection ; we then specialise to sets of _ sr_-matrices and globally optimal solutions .
first of all , we consider the general problem of finding a common subeigenvector for a set of nonnegative matrices ( not necessarily _ sr_-matrices ) .
our results are clearly related to the work in @xcite concerning the intersection of eigencones of commuting matrices over the max and nonnegative algebra . in the next result
, we adopt the notation @xmath201 for @xmath193 and , in an abuse of notation , @xmath202 [ thm : comxsh ] consider the set @xmath203 in ( [ eq : psi ] ) .
the following assertions are equivalent .
* @xmath204 ; * there exists some @xmath205 with @xmath206 for all @xmath164 ; * @xmath207 where @xmath208 for @xmath209 .
( we say that @xmath210 is submultiplicative on @xmath139 ) .
( i)@xmath211(ii ) : first , assume that @xmath204 .
then , there exists @xmath80 such that @xmath212 .
thus , @xmath213 for all @xmath164 .
hence , @xmath206 for all @xmath3 .
( ii)@xmath211(iii ) : suppose that there exists some @xmath80 with @xmath206 for all @xmath164 .
pick some @xmath214 such that @xmath215 where @xmath216 .
then , @xmath217\\\nonumber & \leq \mu(a_{j_p})a_{j_1}\otimes a_{j_2}\otimes\cdots\otimes a_{j_{p-1}}\otimes x\displaybreak[0]\\\nonumber & \vdots\displaybreak[0]\\\nonumber & \leq \mu(a_{j_p})\mu(a_{j_{p-1}})\cdots \mu(a_{j_2 } ) a_{j_{1}}\otimes x\displaybreak[0]\\\nonumber & \leq \mu(a_{j_p})\mu(a_{j_{p-1 } } ) \cdots \mu(a_{j_2 } ) \mu(a_{j_{1 } } ) x.\displaybreak[0]\\\nonumber\end{aligned}\ ] ] writing @xmath218 , we see that @xmath219 from the definition ( [ eq : car ] ) .
hence , @xmath220 . point ( i ) in proposition [ pro : carprop ] implies @xmath221 .
thus , @xmath222 .
note that we essentially used .
( iii)@xmath211(i ) : consider the set of normalised matrices @xmath223 where @xmath224 for all @xmath200 .
pick some @xmath225 . as @xmath210 is submultiplicative on @xmath139 , it is also submultiplicative on @xmath226 .
thus , we have @xmath227 as this is true for any @xmath228 , it follows that @xmath229 taking the @xmath230 root and @xmath231 of both sides , we see that @xmath232 theorem [ thm : mumspsis ] ( ii ) then implies that @xmath233 .
furthermore , since @xmath234 for all @xmath200 we obtain @xmath204 . note that the equivalence ( i)@xmath235(ii ) can be regarded as a special case of hershkowitz - schneider @xcite theorem 2.5 , see also @xcite theorem 3.5 for an extension . in these works ,
the problem of simultaneous nonnegative matrix scaling is considered ; this amounts to finding a diagonal matrix @xmath236 such that @xmath237 for @xmath238 . for our case , take @xmath239 ( the all - ones matrix ) and impose @xmath240 .
however , condition ( iii ) does not appear in @xcite or @xcite . in the terminology of butkovi et al .
@xcite , there exists a * simultaneous visualisation * of all of the matrices in @xmath139 , meaning that @xmath241 for all @xmath3 , and in particular @xmath242 for all @xmath3 and @xmath243 .
the following result for general nonnegative matrices will be useful in the next subsection to clarify the relationship between commutativity and the existence of globally optimal solutions for @xmath244 _ sr_-matrices . [ p : commoncrit ] let @xmath245 have @xmath246 . if @xmath247 , then * @xmath248 for all edges @xmath249 ; * @xmath250 for @xmath251 and @xmath252 . as , @xmath246
, it follows that @xmath253 and @xmath254 .
thus , @xmath255 . from the assumption ,
we obtain @xmath256 it now follows from theorem [ thm : comxsh ] that there exists some @xmath257 with @xmath258 , @xmath259 .
let @xmath260 and consider the diagonally scaled matrices @xmath261 from the choice of @xmath236 it is immediate that @xmath262 furthermore , from well - known facts about diagonal scaling ( see proposition 2.10 of @xcite ) , it follows that @xmath263 and that @xmath264 we prove ( i ) : let @xmath265 be given .
it follows from ( [ eq : ec1 ] ) that @xmath57 is also a critical edge in the digraphs of @xmath266 and @xmath267 .
( [ eq : vis1 ] ) and ( [ eq : mu1 ] ) now imply that @xmath268 hence @xmath269 and this completes the proof .
we prove ( ii ) : let @xmath251 and @xmath252 .
it follows from ( [ eq : ec1 ] ) that @xmath270 is also a critical edge in the digraph of @xmath266 and @xmath271 is a critical edge in the digraph of @xmath267 .
then @xmath272 and hence @xmath273 we next recall the following result , which was established in @xcite and shows that commutativity is a sufficient condition for the existence of a common eigenvector for irreducible matrices .
[ pro : conds ] consider the set @xmath203 in ( [ eq : psi ] ) .
assume that each @xmath274 is irreducible and moreover that @xmath275 then there exists some @xmath276 with @xmath277 for @xmath193 .
the next corollary is an immediate consequence of proposition [ pro : conds ] and the fact , which we recalled in section [ sec : intro ] , that for an irreducible matrix @xmath8 , the set @xmath101 is the subeigencone @xmath95 , which coincides with the eigencone @xmath278 .
[ cor : astar ] consider the set @xmath203 in ( [ eq : psi ] ) .
assume that each @xmath274 is irreducible and moreover that @xmath279 then there exists some @xmath205 with @xmath280 for @xmath193 .
note that implies . in the remainder of this section
, we will only focus on _ sr_-matrices .
we first present the following corollary of theorem [ thm : comxsh ] , which develops the concept of simultaneous visualization for _ sr_-matrices . before stating the corollary , define the * anticritical graph * of an _
sr_-matrix to consist of the edges @xmath281 given by : @xmath282 [ c : comxsh - sr ] consider the set @xmath203 in ( [ eq : psi ] ) .
assume that each @xmath274 is an _ sr_-matrix .
if any of the equivalent statements of theorem [ thm : comxsh ] holds , then there exists some @xmath80 such that for @xmath283 we have @xmath284 in particular , @xmath285 the right - hand side inequality of ( [ eq : sandwich ] ) is the same as theorem [ thm : comxsh ] ( ii ) . for the remaining left - hand side inequality of ( [
eq : sandwich ] ) we observe that @xmath286 is equivalent to @xmath287 .
then we apply @xmath288 .
we next show that two distinct _
sr_-matrices @xmath289 in @xmath290 can not have a common subeigenvector .
let @xmath291 , b=\left [ \begin{array}{cc } 1 & b \\ 1/b & 1 \\ \end{array } \right]\ ] ] and assume that @xmath292 . clearly , @xmath246 and @xmath293 . if @xmath294 , then @xmath295 and @xmath296 . if @xmath297 , then @xmath298 and @xmath299 . in both cases , @xmath300
hence by theorem [ thm : comxsh ] , @xmath8 and @xmath301 do not have a common subeigenvector .
proposition [ pro : conds ] shows that commuting irreducible matrices possess a common max eigenvector .
we now show that for @xmath244 _ sr_-matrices , commutativity is both necessary and sufficient for the existence of a common subeigenvector .
[ rem:3by3set ] for an _
sr_-matrix @xmath302 , it is immediate that all cycle products of length one and two in @xmath52 are equal to 1 .
further , there are two possible cycle products of length 3 in @xmath52 : @xmath303 and @xmath304 . as @xmath8 is an _
sr_-matrix , it follows that @xmath305 and hence at least one of the above products must be greater than or equal to 1 . since @xmath306 , one of the cycles of length three is critical , and the other cycle is anticritical .
thus , @xmath120 and @xmath71 is irreducible .
hence , it follows from proposition [ pro : unique ] that @xmath8 has a unique subeigenvector up to a scalar multiple in @xmath101 which is its max eigenvector .
observe that each edge @xmath270 with @xmath6 belongs either to the critical or to the anticritical graph .
our next result characterises when two _ sr_-matrices in @xmath307 have a common subeigenvector .
[ thm:3b3cond ] let a set @xmath308 of _ sr_-matrices be given . write @xmath309 .
the following are equivalent .
* @xmath204 ; * @xmath8 and @xmath301 commute ; * there exists a vector @xmath198 with @xmath310 , @xmath311 .
the equivalence of ( i ) and ( iii ) follows immediately from theorem [ thm : comxsh ] so we will show that ( i ) and ( ii ) are also equivalent .
\(ii ) @xmath211 ( i ) follows immediately from proposition [ pro : conds ] .
we prove ( iii)@xmath211(ii ) .
first note that it follows from remark [ rem:3by3set ] that for distinct @xmath312 , the edges @xmath57 , @xmath313 are either both critical or both anti - critical for @xmath8 .
the same is true of @xmath301 . calculating @xmath266 and @xmath267 where @xmath260 , it follows from theorem [ thm : comxsh ] , corollary [ c : comxsh - sr ] and the identities @xmath314 , @xmath315 that @xmath316 for any distinct @xmath317 .
it now follows from ( [ eq : olcomm ] ) that for @xmath318 @xmath319\\ \nonumber & = b_{ij}\oplus a_{ij}\oplus b_{ik}a_{kj } \displaybreak[0]\\ \nonumber & = ( b\otimes a)_{ij } \displaybreak[0]\\ \nonumber\end{aligned}\ ] ] where @xmath320 , @xmath321 . rewriting ( [ eq : olcomm ] ) as @xmath322 , it follows readily that @xmath323 and @xmath324 .
it now follows that for @xmath325 , @xmath326\\ \nonumber & = b_{ii}a_{ii}\oplus b_{ij}a_{ji}\oplus b_{ik}a_{ki } \displaybreak[0]\\ \nonumber & = ( b\otimes a)_{ii } \displaybreak[0]\\ \nonumber\end{aligned}\ ] ] thus , @xmath327 as claimed .
it is now straightforward to extend the above result to an arbitrary finite set of _ sr_-matrices in @xmath307 .
[ thm:3by3finite ] let a set @xmath328 of _ sr_-matrices be given .
write @xmath329 .
the following are equivalent .
* @xmath204 ; * @xmath330 for all @xmath331 ; * there exists a vector @xmath198 with @xmath332 for all @xmath3 . as above
, the equivalence of ( i ) and ( iii ) follows immediately from theorem [ thm : comxsh ] and ( ii ) @xmath211 ( i ) follows immediately from proposition [ pro : conds ] . to show that ( i ) @xmath211 ( ii ) , suppose @xmath333 .
then it follows that for all @xmath331 , @xmath334 and hence that @xmath335 . as @xmath336
, it is immediate that @xmath337 for all @xmath331 in @xmath338 .
it follows immediately from theorem [ thm:3b3cond ] that @xmath339 for @xmath340 as claimed .
we note with the following example that commutativity is not a necessary condition for @xmath341 _ sr_-matrices to possess a common subeigenvector .
[ ex : ex4by4 ] consider the _
sr_-matrices given by @xmath342 b=\left [ \begin{array}{cccc } 1 & 4 & 5 & 9 \\ 1/4 & 1 & 1/8 & 9 \\ 1/5 & 8 & 1 & 1/8 \\ 1/9 & 1/9 & 8 & 1 \\ \end{array } \right ] \ ] ] where @xmath204 . here ,
@xmath343^t$ ] is a common subeigenvector .
however , it can be readily verified that @xmath344 .
in general , it will not be possible to find a single vector @xmath112 that is globally optimal for the set @xmath139 of _ sr_-matrices given by ( [ eq : psi ] ) . with this in mind ,
in this short section we consider a different notion of optimal solution for the multiple objective functions @xmath191 , @xmath193 .
in fact , we consider the following optimisation problem .
@xmath345 in words , we are seeking a weight vector that minimises the maximal relative error where the maximum is taken over the @xmath166 criteria ( _ sr_-matrices ) .
corollary [ cor : cprprop ] has the following interpretation in terms of the optimisation problem given in ( [ eq : opt2 ] ) .
[ pro : opt2 ] consider the set @xmath139 given by ( [ eq : psi ] ) . then : * @xmath346 ; * @xmath112 solves ( [ eq : opt2 ] ) if and only if @xmath347 .
corollary [ cor : cprprop ] shows that there exists some @xmath198 with @xmath348 if and only if @xmath349 .
( i ) follows from this observation .
the result of ( ii ) is then immediate from the definition of @xmath155 .
0.2 cm thus far , we have considered two different approaches to the multi - objective optimisation problem associated with the ahp . in this section we turn our attention to what is arguably the most common framework adopted in multi - objective optimisation : pareto optimality @xcite . as
above , we are concerned with the existence of optimal points for the set of objective functions @xmath192 , for @xmath193 associated with the set @xmath139 ( [ eq : psi ] ) of _ sr_-matrices .
we first recall the notion of _ weak pareto optimality_. [ def : wpop ] @xmath350 is said to be a _ weak pareto optimal point _ for the functions @xmath351 if there does not exist @xmath276 such that @xmath352 for all @xmath353 .
the next lemma shows that every point in the set @xmath155 is a weak pareto optimal point for @xmath354 .
[ lem : wpop ] let @xmath160 be given by ( [ eq : psi ] ) . any @xmath355 is a weak pareto optimal point for @xmath354 .
let @xmath356 be given .
then @xmath357 for @xmath358 .
if there exists some @xmath198 such that @xmath359 for @xmath193 , then for this @xmath112 @xmath360 for @xmath193 .
this contradicts proposition [ pro : opt2 ] .
we next recall the usual definition of a pareto optimal point .
[ def : pop ] @xmath350 is said to be a _
pareto optimal point _ for the functions @xmath361 if @xmath362 for @xmath193 implies @xmath363 for all @xmath193 .
we later show that the multi - objective optimisation problem associated with the ahp always admits a pareto optimal point .
we first present some simple facts concerning such points . [ thm : pops ] let @xmath160 be given by ( [ eq : psi ] ) .
then : * if @xmath355 is unique up to a scalar multiple , then it is a pareto optimal point for @xmath354 ; * if @xmath364 is unique up to a scalar multiple for some @xmath200 , then it is a pareto optimal point for @xmath354 .
observe that both conditions imply @xmath365 .
\(i ) assume that @xmath355 is unique up to a scalar multiple .
pick some @xmath80 such that @xmath362 for all @xmath3 . then , @xmath366 for all @xmath3 which implies that @xmath367 .
thus , @xmath368 for some @xmath369 .
hence , @xmath363 for all @xmath3 and @xmath15 is a pareto optimal point .
\(ii ) assume that for some @xmath164 , @xmath364 is unique up to a scalar multiple .
suppose @xmath80 is such that @xmath370 for all @xmath371 .
in particular , @xmath372 , and this implies that @xmath368 for some @xmath373 .
further , it is immediate that for any other @xmath374 ( @xmath6 ) , we have @xmath375 .
thus , @xmath15 is a pareto optimal point . by proposition [ pro : unique ] ,
condition ( i ) is equivalent to @xmath187 and @xmath186 to be irreducible , and condition ( ii ) is equivalent to @xmath376 and @xmath377 to be irreducible for some @xmath3 .
[ cor : pops ] let the set @xmath378 given by ( [ eq : psi ] ) consist of _ sr_-matrices . for @xmath379 ,
any @xmath380 is a pareto optimal point for @xmath354 .
notice that from remark [ rem:3by3set ] for @xmath381 case , there exists a unique subeigenvector up to a scalar multiple in each @xmath196 for @xmath173 .
this is also true for the @xmath382 case because @xmath120 and @xmath71 is irreducible .
the result directly follows from ( ii ) in theorem [ thm : pops ] .
the following example demonstrates point ( i ) in theorem [ thm : pops ] .
[ ex : popsdpu ] consider the following matrices given by @xmath383 b=\left [ \begin{array}{cccc } 1 & 1/2 & 4 & 1/8 \\ 2 & 1 & 3 & 2 \\ 1/4 & 1/3 & 1 & 5 \\ 8 & 1/2 & 1/5 & 1 \\ \end{array } \right ] .\ ] ] @xmath144 matrix is obtained as follows @xmath384\ ] ] where @xmath187 and @xmath186 is irreducible .
from proposition [ pro : unique ] , we have a unique vector ( up to a scalar multiple ) in @xmath155 : @xmath385^t$ ] .
figure [ fig : ex1 ] below represents the values of @xmath386 and @xmath387 at @xmath15 and some points in @xmath388 , @xmath389 and @xmath43 .
remark that pareto optimality is observed at @xmath390 where @xmath391 .
, scaledwidth=80.0% ] our objective in the remainder of this section is to show that the multi - objective optimisation problem associated with the ahp _ always _ admits a pareto optimal solution .
we first recall the following general result giving a sufficient condition for the existence of a pareto optimal point with respect to a set @xmath392 . essentially , this is a direct application of the fact that a continuous function on a compact set always attains its minimum .
[ thm : existence ] let @xmath393 be nonempty and compact .
let a set of continuous functions @xmath394 be given where @xmath395 for @xmath193 .
there exists @xmath396 such that @xmath397 , @xmath398 for @xmath193 implies @xmath399 for @xmath193 .
this result follows from elementary real analysis and the observation that if @xmath15 minimises the ( continuous ) weighted sum @xmath400 where @xmath401 for @xmath193 , then @xmath15 must be pareto optimal for the functions @xmath402 . a point @xmath15 satisfying the conclusion of the theorem [ thm : existence ] is said to be pareto optimal for @xmath403 _ with respect to _ @xmath404 .
thus , for any multi - objective optimisation problem with continuous objective functions defined on a compact set , there exists a point that is pareto optimal with respect to the given set . to apply the above result to the ahp
, we first note that for a set @xmath139 of _
sr_-matrices , @xmath157 is compact by theorem [ thm : geop ] ( recall that _ sr_-matrices are positive and hence irreducible ) .
we now show that any point in @xmath157 that is pareto optimal with respect to @xmath157 is also pareto optimal with respect to the set @xmath43 .
although we do nt specifically use the _ sr _ property in the following , we assume @xmath139 to consist of _ sr_-matrices as we are primarily interested in the ahp application . instead , we could just assume that @xmath150 are irreducible implying that @xmath157 is compact and a pareto optimal point exists .
[ lem : existence ] consider the set @xmath139 in ( [ eq : psi ] ) and assume that @xmath150 is an _
sr_-matrix for @xmath193 .
let @xmath15 be a pareto optimal point for @xmath405 with respect to @xmath157 . then , @xmath15 is also pareto optimal for @xmath354 with respect to @xmath43 .
assume that @xmath406 is a pareto optimal point with respect to @xmath157 .
suppose @xmath407 .
then from the definition of @xmath155 ( [ eq : cpr ] ) , it follows that @xmath408 as @xmath409 , @xmath357 for @xmath193 .
it follows immediately from ( [ eq : par1 ] ) that for any @xmath410 , it can not happen that @xmath411 for @xmath193 .
let @xmath112 in @xmath155 be such that @xmath412 as @xmath413 for all @xmath414 , @xmath193 , and @xmath15 is pareto optimal with respect to @xmath157 , it follows that @xmath415 for @xmath193 . our next step is to show that there exists a point @xmath416 that is pareto optimal with respect to @xmath157 .
[ pro : existence ] consider the set @xmath139 in ( [ eq : psi ] ) and assume that @xmath150 is an _
sr_-matrix for @xmath193 .
there exists @xmath416 that is pareto optimal for @xmath417 with respect to @xmath157 .
first note that @xmath418 since @xmath419 .
theorem [ thm : geop ] shows that @xmath157 is compact .
furthermore , for any irreducible matrix @xmath8 , the function @xmath420 is a composition of continuous functions and hence continuous on a compact set .
theorem [ thm : existence ] implies that there exists @xmath15 in @xmath157 that is pareto optimal with respect to @xmath157 .
combining proposition [ pro : existence ] with lemma [ lem : existence ] , we immediately obtain the following result .
[ cor : existence ] consider the set @xmath139 in ( [ eq : psi ] ) and assume that @xmath150 is an _
sr_-matrix for @xmath193 .
there exists @xmath416 that is pareto optimal for @xmath417 with respect to @xmath43 .
corollary [ cor : existence ] means that there exists a vector @xmath112 of positive weights that is simultaneously pareto optimal and also optimal in the _ min - max _ sense of section [ sec : minmax ] for the error functions @xmath192 , @xmath193 . finally , to illustrate the above results , we revisit example [ ex : saaty ] . [ ex : saaty2 ] let @xmath421 be as in example [ ex : saaty ] .
taking @xmath422 , @xmath423 to be the @xmath3th entry of the max eigenvector of @xmath20 , normalised so that @xmath424 , we apply theorem [ thm : existence ] to compute pareto optimal solutions in the set @xmath157 by minimising the weighted sum @xmath425 using the matlab function _ fminsearch_. observe that @xmath426 , so there is no common subeigenvector in this case .
next , we calculate the max eigenvector of @xmath427 @xmath428\ ] ] we find that there are multiple pareto optimal points giving at least two possible distinct rankings : @xmath38 and @xmath429 .
notice that first ranking scheme is the same as the one obtained from the classical method used in example [ ex : saaty ] .
the second ranking scheme is also reasonable , since if we analyse the local rankings associated with the set of _ sr_-matrices in detail , we see that @xmath430 for @xmath26 , @xmath431 and @xmath432 .
in particular , @xmath30 is preferred to all other alternatives for @xmath431 .
building on the work of elsner and van den driessche @xcite , we have considered a max - algebraic approach to the multi - criteria ahp within the framework of multi - objective optimisation .
papers @xcite characterise the max eigenvectors and subeigenvectors of a _ single _ _ sr_-matrix as solving an optimisation problem with a single objective .
we have extended this work to the multi - criteria ahp by directly considering several natural extensions of this basic optimisation problem to the multiple objective case .
specifically , we have presented results concerning the existence of : globally optimal solutions ; min - max optimal solutions ; pareto optimal solutions .
the principal contribution of the paper is to draw attention to this max - algebraic perspective on the multi - criteria ahp , with the main results in this direction being : establishing the connection between the generalised spectral radius and min - max optimal solutions ( proposition [ pro : opt2 ] ) ; proving the existence of pareto optimal solutions and showing that it is possible to simultaneously solve the pareto and min - max optimisation problems ( proposition [ pro : existence ] and corollary [ cor : existence ] ) .
we have also related the existence of globally optimal solutions to the existence of common subeigenvectors and highlighted connections between this question and commutativity ( theorem [ thm:3b3cond ] ) .
buket benek gursoy and oliver mason acknowledge the support of irish higher educational authority ( hea ) prtli network mathematics grant ; serge sergeev is supported by epsrc grant rrah15735 , rfbr - cnrs grant 11 - 01 - 93106 and rfbr grant 12 - 01 - 00886 .
b. heidergott , g. j. olsder , j. van der wounde , max plus at work : modeling and analysis of synchronized systems , a course on max - plus algebra and its applications , princeton university press , princeton , nj ( 2006 ) .
d. hershkowitz , h. schneider , one - sided simultaneous inequalities and sandwich theorems for diagonal similarity and diagonal equivalence of nonnegative matrices , elec .
j. of linear algebra 10 ( 2003 ) 81 - 101 .
krivulin , evaluation of bounds on the mean rate of growth of the state vector of a linear dynamical stochastic system in idempotent algebra , vestnik st .
petersburg university : mathematics 38 ( 2005 ) 45 - 54 . | the analytic hierarchy process ( ahp ) is widely used for decision making involving multiple criteria . elsner and van den driessche
@xcite introduced a max - algebraic approach to the single criterion ahp .
we extend this to the multi - criteria ahp , by considering multi - objective generalisations of the single objective optimisation problem solved in these earlier papers .
we relate the existence of globally optimal solutions to the commutativity properties of the associated matrices ; we relate min - max optimal solutions to the generalised spectral radius ; and we prove that pareto optimal solutions are guaranteed to exist .
+ _ keywords : _ analytic hierarchy process ( ahp ) , _
sr_-matrix , max algebra , subeigenvector , generalised spectral radius , multi - objective optimization .
+ _ ams codes : _
91b06 , 15a80 , 90c29 |
over the last two decades , the rapid development of single molecule manipulation techniques has produced dramatic advances in chemistry and molecular physics .
the fields of dna biotechnology and nanofabrication in particular , have benefited from these advances . in turn
, successes in the field of dna biotechnology have further motivated the study of dna manipulation techniques , with a strong focus on studying mechanisms of inducing dna dehybridization ( @xcite , @xcite ) . in this study , we consider dehybridizing dna by pulling dna molecules through a transmembrane channel . this work is motivated in part by the experimental results of nakane _
@xcite , and bates _ el al .
_ @xcite . in the experiments of nakane , a probe dna that extends 14 base pairs beyond the opening of a trans - membrane channel
is used to capture a 14 base pair target strand ( see fig [ fig1 ] ) . in order to prevent complete passage of the probe dna through the membrane channel
, the probe was attached to an avidin anchor protein by a 50-base poly - a tail .
the target strands used in the experiments were either completely homologous to the probe strand , or had one nonhomologous defect " base .
after the probe strand had captured a target strand , a transmembrane voltage was applied to pull the probe strand through an @xmath0-hemolysin channel embedded in a lipid membrane . as the @xmath0-hemolysin channel could only accommodate _
single - stranded _ dna molecules , any target strand hybridized with the probe dna would need to detach completely before the probe could be fully drawn into the channel .
as the transmembrane voltage was applied , the ionic current through the @xmath0-hemolysin channel was simultaneously monitored . when any part of the probe strand was inside the channel , the channel was blocked , and the ionic current fell below the measurement threshold .
when the probe strand was completely pulled through the channel ( after the target strand detached ) , ionic currents resumed and were measured .
thus , in the experiment of nakane , the distribution of first passage times , @xmath1 , for the escape of the probe dna from the @xmath0-hemolysin channel was measured . in order to understand the first passage times observed in these experiments , we have produced estimates for the thermally averaged mean first passage time @xmath2 using simple two - dimensional and one - dimensional stochastic models .
we model the dna hybridization energetics using both a simple free energy model where each hybridized base contributes an equal amount to the total free energy , and using free energies produced by the mfold 2-state hybridization server @xcite .
our predicted mean first passage times show the same qualitative features observed by nakane _
_ under the influence of small transmembrane voltages , the first passage times are found to depend strongly on the presence and energy of defects in the target dna strand . as the transmembrane voltage is increased , a distinct roll - over in the predicted first passage times
is observed , and the first passage times become relatively insensitive to the presence of defects on the target dna strand . we propose that this roll - over in @xmath3 represents a transition between a thermally - dominated dehybridization mechanism in the low voltage regime , and a driven stripping mechanism in the high voltage regime .
target dna is assumed to occur via two coordinates @xmath4 and @xmath5 , representing unzipping from the two ends of the probe dna . here
, an @xmath6 target strand is shown with a single defect .
the probe dna is shown in blue , hybridized target dna in green , the defect base in red , dehybridized bases in brown , and extracted bases in light blue .
the probability of bubbles forming is ignored .
, width=316 ] motivated by the double - stranded dna unzipping models of cocco _ et .
_ @xcite and poland and scheraga @xcite , we model the process of dna probe extraction using a two - sided zipper model described in fig .
this model generates @xmath7 distinct partial dehybridization states labeled @xmath8 , were @xmath9 is the number of bases at the @xmath10 end of the probe that are dehybridized _ and _ have been pulled into the transmembrane region , and @xmath11 is the number of dehybridized bases at the @xmath12 end of the dna probe . using this nomenclature , the dna probe is fully dehybridized when @xmath13 , and fully _ extracted _ only when @xmath14 and @xmath15 .
note that his model neglects the effects of internal dehybridization bubbles , which will be discussed later in this section . defining @xmath16
to be the probability of being in state @xmath8 at time @xmath17 , given that the system was initially in state @xmath18 , we find the system evolves according to a master equation @xmath19 in order to produce the elements @xmath20 of the transition matrix , we define the following five transition mechanisms between states @xmath8 : @xmath21 bond breaking : : [ rate3 ] the hydrogen bonds between bases at the @xmath10 end of the probe dna break , the probe dna base moves into the transmembrane region , and the state shifts from @xmath8 to @xmath22 .
this transition path includes a bond breaking event , immediately followed by translation of the probe dna , and occurs at rate + @xmath23 + where @xmath24 is the change in free energy between the current state @xmath25 and the target state @xmath26 , and @xmath27 is a phenomenological attempt frequency .
we note that we could express @xmath27 as the product of an attempt frequency and a free energy barrier @xmath28 .
however for the @xmath10 bond - breaking transition , as well as all other transitions in our model , we pull any dependence on the barrier energy into the effective attempt frequency . as this state transition involves a net translation of a partially charged nucleotide across the membrane , a drop in the interaction energy between the applied electric field and the probe dna occurs : @xmath29 + where @xmath30 is the mean partial charge of a dna base . @xmath21
bond formation : : [ rate4 ] in the reverse of the @xmath10 bond breaking transition , the probe dna translates one base in the @xmath12 direction , and corresponding bases on the probe and the target dna strands hybridize .
this shifts the system from the @xmath8 state to the @xmath31 state , and occurs at rate + @xmath32 @xmath33 bond breaking : : [ rate1 ] the hydrogen bonds between two hybridized bases located at the @xmath12 end of the probe break , thus increasing @xmath11 by one .
this transition is assumed to occur with rate + @xmath34 @xmath35 bond formation : : [ rate2 ] corresponding dehybridized bases at the @xmath12 end of the probe strand and the @xmath10 end of the target dna strand form a bond , decreasing @xmath11 by one .
this transition is assumed to occur at rate + @xmath36 @xmath37 free translation : : [ rate5 ] in the case where the dna probe is completely dehybridized @xmath38 , but not fully extracted @xmath39 , the dna probe will translate under the influence of the applied potential .
the translation changes the state from @xmath8 to @xmath40 or @xmath41 and occurs with rate + @xmath42 + the two rates @xmath43 and @xmath44 describe translation in the @xmath10 and @xmath12 directions respectively .
the only significant difference between @xmath43 and @xmath44 is the sign of the change in the interaction energy between the probe dna and the applied electric field . in the case of @xmath10 translation @xmath45 , while in the case of @xmath12 translation @xmath46 .
having defined the possible transitions in the model , we are left with five unknown attempt frequencies @xmath47 , and an unknown value for the mean partial charge per dna base @xmath30 . based on the results of nakane , we set @xmath48 @xcite , were @xmath49 is the electron charge .
since free translation and @xmath10 bond breaking both involve sliding of the probe dna strand , the time scale for these transitions is set approximately by the time scale for translation of the single - stranded probe dna . to find this time scale
, we built a simplified model of single - stranded dna translation consisting of only the free translation moves described above . using this model
, we simulated the complete passage of a 60-base single - stranded dna segment through a channel 30 dna bases in length .
we then compared the mean first passage time for complete transit of single - stranded dna computed from this model , with the experimental results of bates _
@xcite for dna transport through @xmath0-hemolysin channels at low voltages ( @xmath50 at 20mv , @xmath51 at 40mv ) .
the best match between the computed and experimental results occurred when the attempt frequency was approximately @xmath52 . to determine a value for @xmath53 , we followed cocco _
_ @xcite and set @xmath53 to equal the inverse self diffusion time for a single nucleotide ( @xmath54 ) . with rough order of magnitude estimates for @xmath27 and @xmath53
we then tuned the parameters of our model to give a reasonable match to the experimental data of nakane @xcite for the variation of escape time with voltage .
after hand tuning , the parameters were @xmath55 , @xmath56 , @xmath57 , @xmath58 , and @xmath59 . while easy to analyze , our model fails to capture a number of aspects of the denaturation and extraction process .
for example , by assuming the short probe - target dna complex denatures only from the ends , we have ignored the effects of bubble formation . by bubbles we refer to dehybridized regions of dna that are bounded by at least one hybridized base on both sides . while clearly an approximation , we do not believe that bubble formation will qualitatively affect the process of driven dna hybridization described here . under physiologic conditions , structural fluctuations in double stranded dna lead to the formation of bubbles that are typically on the order of several tens of base pairs @xcite .
in this paper we consider the dehybridization and extraction of dna strands only 10 - 20 bases in length , and thus considerably smaller than the typical bubble size found in dna at physiologic conditions .
additionally , recent experiments @xcite have studied the formation of bubbles in small , atypically bubble - prone , poly - at dna chains approximately @xmath60 bases in length . here ,
the formation of dna bubbles approximately @xmath61 bases in length were observed in the dna strand .
however , through fluorescence - correlation spectroscopy altan - bonnet @xcite were able to measure the typical lifetime of these small bubbles to be approximately 50@xmath62 .
this lifetime is typically much smaller than the mean time for the dehybridization and extraction of small dna strands through @xmath0-hemolysin at moderate driving voltages .
additionally we do not model partial - registry binding , where the target dna binds to the probe dna with an offset so that the target dna `` over - hangs '' the probe .
if we assume that the target - probe dna complex is well equilibrated prior to the start of the extraction processes , then such out of registry binding is unlikely due to its high energetic cost .
moreover , should such binding occur , its effects on the dna dehybridization process would not be qualitatively different from the effects of starting in a @xmath8 state with a similar number of unhybridized bases . finally , we do not model the full extraction of the probe dna ( _ i.e. , _ full removal from the @xmath0-hemolysin channel ) . as re - binding of the target dna is unlikely
once the probe is fully drawn into the channel ( typical target concentrations are 10@xmath63 m @xcite ) , complete removal of the probe dna depends primarily on the details of the dna - channel interactions .
previous studies @xcite , have shown that translation of single - stranded dna 60 bases in length under moderate driving ( 0.04v ) in an @xmath0-hemolysin channel occurs on a time scale of 500@xmath63s . rather than model the interactions between the probe dna and channel
, we can simply add @xmath64 to the first passage times for probe escape we compute using our model . as we shall see , at low to moderate membrane voltages this additional time is small compared to the time required to denature the probe dna .
we use eqn [ eqn1 ] to compute the thermally averaged time for the dna probe to first reach the fully extracted state @xmath65 . to do this
, we make the @xmath66 state completely absorbing , and slightly modify eqn [ eqn1 ] by setting @xmath67 for all @xmath17 : @xmath68 note that the first term in eqn [ fpt0 ] does not include contributions from transitions out of state @xmath66 while the second term does include transitions into state @xmath66 . upon defining @xmath69 to be the probability that the system reaches the absorbing state between time @xmath17 and time @xmath70 , given that it started in state @xmath71 we find @xmath72 with @xmath69 computed from the solution to eqn [ fpt0 ] , we can determine the mean first passage time to the extracted state @xmath73 finally , to find the thermally averaged mean first passage time , @xmath3 , we sum the results of eqn [ fpt2 ] over all possible initial states of the probe dna , @xmath74 where @xmath75 is the free energy of state @xmath8 relative to the free energy of the completely dehybridized system .
the validity of eqn [ fpt3 ] is contingent on the assumption of local thermal equilibrium between the probe and the fluid bath .
however , as the probe is typically maintained in the inserted state for several tenths of a second prior to removal @xcite , we expect this assumption to hold to reasonable accuracy . ) . at low voltages , energetic effects account for the majority of the difference in escape times between the various defect - carrying strands .
inset : thermally averaged mean escape times of a dna segment 5 bases long .
each hybridized base lowers the free energy by @xmath76 , except for a single defect site ( located at position @xmath77 ) which lowers the free energy by @xmath78 when hybridized .
placing the defect in the center of the segment lowers the mean escape time significantly more than placing a defect at either end of the segment.,width=316 ] to investigate how nonhomologous defects influence the mean first passage time , we first analyze the process of dna extraction using a simple model where each non - defective hybridized base pair lowered the free energy of the probe - target complex by a constant amount . a single defective base pair with a binding free - energy lower than all other bases
was then introduced into the target strand . as an example
, we considered a short strand of eleven bases , and used the double zipper model to generate the mean escape time curves displayed in the inset of fig [ lifetime1 ] .
there are two features worth noting ; the mean escape time clearly depends on the location of a defective base as well as the energy of the defect , and the position and energy of the defect influence the escape time less as the applied voltage is increased . at first glance
the dependence of the escape time on defect location appears counter - intuitive given that the defect energy did not vary with the location of the defect along the target strand . nonetheless , the behavior displayed in fig .
[ lifetime1 ] where a defect placed in the center of the dna strand produces maximal reduction in mean escape time , is extremely general and rather insensitive to changes in the parameter values of the model . to understand the source of this effect , consider a dna chain with a defect at position @xmath77 , when the dna chain is in the @xmath79 or @xmath80 state with @xmath81 .
if the defect is at site @xmath77 , the probability of the adjacent dna base rehybridizing goes as @xmath82 . here
@xmath83 is the rate at which the defect base dehybridizes , and @xmath84 is the rate of rehybridization of the base adjacent to the defect .
the effect of introducing a defect is to increase @xmath83 while keeping @xmath84 constant , reducing the likelihood of hybridization of the adjacent base . in the case where the defect results in particularly weak binding ,
the increase in @xmath83 can be quite large , making it highly unlikely that dna will rehybridize beyond the defect base @xmath77 once the defect itself has dehybridized .
this shortens the effective length of the dna by @xmath85 sites .
we note however that introducing bubbles into the model would minimize the variation in escape time with defect position . in the presence of bubbles ,
bases well separated from the defect could rehybridize without the defect base itself rehybridizing .
this would dramatically reduce the variation in the mean escape time associated with the defect position .
nonetheless , the mean escape time of gc - rich strands which are relatively resistant to bubble formation may still be sensitive to the position of defects in the target strand .
.dna target sequences @xcite along with the hybridization energies predicted by mfold ( @xmath86 ) .
the probe dna strand is perfectly complementary to the 14pc sequence .
the difference in the free energies between the various defect strands is due to the presence of different defect nucleotides ( * a * vs * c * ) , and differences in the bases neighboring the defects . for more detail
see @xcite . [ cols="^,^,^",options="header " , ] turning to a more realistic free energy model , we used the the mfold 2-state hybridization server @xcite to generate free energy surfaces for four distinct probe sequences , each 14 bases in length @xcite ( see table [ sequences ] ) .
the form of the lifetime verses voltage behavior produced by eqn [ fpt3 ] for these sequences is displayed in fig .
[ lifetime1 ] .
the four voltage response curves clearly separate into two groups ; the 14pc sequence ( perfectly complementary to the dna probe ) and the 1a sequence ( with a single non - complementary base at the @xmath10 end of the sequence ) belong to a long lived set , while the 7c and 10c sequences ( both with a single non - complimentary base in the center of the sequence ) form a group with a dramatically shorter lifetime .
we note that the approximately four orders of magnitude difference between the defect - free and 1a escape times and the 7c and 10c escape times at zero voltage , is consistent with the free energy differences between the various probes . from the standpoint of potential applications to dna sequencing ,
the reduction in free energy produced by a defective base is clearly the dominant influence on the mean escape time .
returning to the dna sequences used to produce fig [ lifetime1 ] , we investigated the consequences on the mean probe lifetime of varying the @xmath12 detachment rate , and found that this rate has a significantly greater influence on the mean escape time in the low voltage regime than in the high voltage regime .
this leads us to propose that the field - induced dehybridization can occur through one of two mechanisms . in the low voltage regime , thermal denaturation of the target dna from the probe
is the dominant mode of separation . in this region of the lifetime - voltage curve , thermally - induced dehybridization from the @xmath12 end is at least as significant as voltage driven - bond stripping at the @xmath10 end .
as the applied voltage is increased , the system enters a transition region where voltage - driven bond stripping becomes the statistically favored mechanism of detachment .
finally , in the high voltage region stripping becomes the primary means of separation .
to confirm this hypothesis , we find the probability that base @xmath77 ( measured from the @xmath10 end of the probe dna ) , is the last base to dehybridize prior to extraction of the dna probe .
we first define @xmath87 to be the probability density that the last dehybridization event occurs at base @xmath77 between time @xmath17 and @xmath70 . making use of the solution to eqn [ eqn1 ]
, @xmath88 can be expressed as @xmath89 here @xmath90 is the probability the probe dna is fully drawn into the channel by time @xmath91 given that it was fully dehybridized ( _ i.e. , _ in state @xmath92 ) at time @xmath17 , assuming that no rehybridization events occur between @xmath17 and @xmath91 . rates @xmath93 and @xmath94 are defined by eqns [ bondbreakrate1 ] and [ bondbreakrate2 ] respectively . since the probabilities produced by eqn [ eqn1 ] contain contributions from paths through the @xmath8 state space that include rehybridization events , we can not use the solution to eqn [ eqn1 ] to compute @xmath95 .
however we can compute @xmath96 by considering the master equation governing the transitions between fully dehybridized states @xmath97 . defining @xmath98
to be the probability of being in the @xmath99 state at time @xmath17 , given that the system was in state @xmath100 at time @xmath101 , we have @xmath102 \displaystyle \frac{dw_{(i , n - i)}}{dt } & = & -(r_2^{(i ) } + r_1^{(i ) } + k_3^{(i ) } + k_4^{(i)})w_{(i , n - i ) } & \\[13pt ] & & + k_3^{(i-1)}w_{(i-1,n-(i-1 ) ) } + k_4^{(i+1)}w_{(i+1,n-(i+1 ) ) } & \\[13pt ] & & 0 < i < ( n-1 ) & \\[13pt ] \displaystyle \frac{dw_{(n-1,1)}}{dt } & = & -(r_2^{(n-1 ) } + r_1^{(n-1 ) } + k_3^{(n-1 ) } + k_4^{(n-1)})w_{(n-1,1)}\\[13pt ] \displaystyle & & + k_3^{(n-2)}w_{(n-2,2)}. & \\[13pt ] \end{array } \label{auxmaster1}\ ] ] .
the configurational free energies were produced assuming a simple free energy model with a uniform free energy per bond , and uniform attempt frequencies of @xmath103.,width=316 ] in eqn [ auxmaster1 ] there are two distinct sets of absorbing states ; the first set contains only the fully extracted state @xmath66 , the second set contains the states with only one hybridized base @xmath104 . as we are seeking the probability of reaching the fully extracted state without having any base rehybridize , we are only concerned with the @xmath66 absorbing state .
the probability distribution for escape from the set of fully dehybridized states @xmath99 to the fully extracted state @xmath66 is @xmath105 .
then @xmath90 is @xmath106 & = & k_3^{(n-1 ) } \int_0^{t^ { * } - t } w_{(n-1,1)}(t;(i , n - i ) ) dt.\\[13pt ] \end{array } \label{auxmaster2}\ ] ] substituting eqn [ auxmaster2 ] into eqn [ pathprob0 ] , and letting @xmath107 , we have @xmath108 & & + p_{i}(t ) r_{1}^{(i ) } k_3^{(n-1 ) } \int_0^{\infty } w_{(n-1,1)}(t;(i , n - i ) ) dt.\\[13pt ] \end{array } \label{pathprob1}\ ] ] note that in the limit of low target dna concentrations , once the probe and target dna have completely dehybridized , rehybridization is unlikely during the time required to extract the probe dna .
thus , in the case of low target concentrations , the integral @xmath109 and eqn [ pathprob1 ] yields @xmath110 . with @xmath88 , we can compute the probability , @xmath111 , that the last base to dehybridize is base @xmath77 from the expression @xmath112 .
the thermally averaged probability that base @xmath77 is the last to dehybridize is computed from @xmath113 referring to the inset in fig [ bondbreak1 ] , we see that using the simple free energy model where each hybridized base lowers the free energy of the system by @xmath114 produces a @xmath115 that is symmetric about the mid - point of the dna lattice .
this is exactly what is expected in the case where all the attempt frequencies in the model are equal and thermal denaturation dominates the dehybridization process . as the voltage increases ,
the behavior of @xmath115 is also predictable , with the bases at the @xmath12 end of probe dna becoming more likely to be the last to separate .
this is consistent with a voltage - driven denaturation mechanism where stripping of the target dna occurs . using the mfold free - energies for the 10c dna segment , we see a similar effect where the free energy surface dominates the last bond - probability at low voltages . at higher voltages the effects of the free energy surface
are obscured and a stripping mechanism dominates .
this is consistent with fig [ lifetime1 ] , where we see that the mean first passage times are most strongly influenced by sequence defects in the low voltage limit .
this transition between a thermally dominated denaturation mechanism and a direct stripping mechanism can explain the decreased influence of defects on the mean extraction time at high membrane voltages . to demonstrate this
, we consider the one - dimensional reduction of the two - dimensional model of dna dehybridization already presented .
retaining only the @xmath10 hybridization transitions , the master equation for the one - dimensional model is @xmath116 & & k_1(n_1 - 1)(1-\delta_{n_1,0})p(n_1 -1,0 ) + & \\[13pt ] & & r_1(n_1 + 1)(1-\delta_{n_1,n-1})p(n_1 + 1,0 ) , & \\[13pt ] & & \mathrm { } 0 \leq n_1 < n-1 . &
\\[13pt ] \end{array } \label{master1}\ ] ] the one - dimensional model of dna denaturation described by eqn [ master1 ] is equivalent to a one dimensional random walk with a reflecting boundary at @xmath117 and an absorbing boundary at @xmath14 .
the first passage time for extraction into the transmembrane channel is then @xcite @xmath118 and the thermally averaged escape time is @xmath119
to make contact with the two dimensional model , we take the high voltage limit of eqn [ fprw1 ] .
note that in the high voltage limit we expect the extraction times predicted by the one and two dimensional models to agree to high accuracy .
additionally , we focus on @xmath120 as this is by far the dominant term in eqn [ fprw2 ] . in the limit of high voltage @xmath121 and @xmath122 here @xmath123 .
eqn [ fprw3 ] shows that in the high voltage limit , the mean extraction time is closely approximated by the time required to independently dehybridize @xmath124 dna bases without significant rehybridization occuring . in this limit
, a defect only influences the extraction process once , when the defect base pair is first dehybridized .
conversely , if we consider the zero voltage limit , eqn [ fprw1 ] yields @xmath125 where the second term includes contributions from rehybridization events .
the effect of a defect is to reduce the ratio @xmath126 , and thus at low voltages a defect accelerates the extraction process each time the defect base pair is denatured . at low voltages
the base pair is typically denatured multiple times prior to probe exaction , and hence defects have a more pronounced effect on the mean extraction time at low voltages than they do at higher voltages .
we also note that the magnitude of eqn [ fprw4 ] depends strongly on the position of the defect in the dna sequence .
as was the case in the two - dimensional model , the mean extraction time predicted by the one - dimensional model is a minimum when the defect is located in the middle of the target strand .
as before , we ascribe this position dependence to a defect s ability to reduce the likelihood of rehybridization past the defect base pair , thus shortening the effective length of the probe - target complex .
we have shown that the dehybridization process critical for the voltage - driven transport of dna through membrane channels can proceed through two distinct mechanisms . in the low voltage regime ,
thermal dehybridization is the dominant mechanism and the details of the hybridized dna s free energy surface set the time scale for transport . in the moderate to
high voltage regime , direct stripping of the complementary dna off the probe dna is the dominant dehybridization mechanism . driven primarily by the applied voltage ,
this dehybridization mechanism is comparatively insensitive to the shape of the hybridized dna s free energy surface .
this behavior may have implications for the use of dna sequencing devices based on nanopores .
specifically , an attempt to increase the sequencing rate by increasing the voltage driving the double stranded dna through the nanopore , may obscure the distinction between different dna sequences .
this may limit the maximum rate at which sequencing can be performed .
gl and gnp acknowledge financial support from the natural science and engineering research council of canada .
gl and tc acknowledge support from the national science foundation through grant dms-0206733 , and the national institutes of health through grant k25 ai058672 .
the authors thank andre marziali , john nakane , and matthew wiggin for valuable comments on the manuscript . | we investigate the voltage - driven transport of hybridized dna through membrane channels . as
membrane channels are typically too narrow to accommodate hybridized dna , the dehybridization of the dna is the critical rate limiting step in the transport process . using a two - dimensional stochastic model ,
we show that the dehybridization process proceeds by two distinct mechanisms ; thermal denaturation in the limit of low driving voltage , and direct stripping in the high to moderate voltage regime .
additionally , we investigate the effects of introducing non - homologous defects into the dna strand . |
the study of the final state of the gravitational collapse of initially regular distributions of matter is one of the open problems in classical general relativity , having attracted remarkable attention in past decades .
when a sufficiently massive star exhausts all the thermonuclear sources of its energy , it would undergo a collapsing scenario due to its own gravity , without reaching a final state in terms of a neutron star or white dwarf . under a variety of circumstances , singularities will inevitably emerge ( geodesic incompleteness in space - time ) , matter densities and space - time curvatures diverging .
albeit the singularity theorems @xcite state that there exist space - time singularities in a generic gravitational collapse , they provide no information on the nature of singularities : the problem of whether these regions are hidden by a space - time event horizon or can actually be observed , remains unsolved .
the cosmic censorship conjecture ( ccc ) , as hypothesized by penrose @xcite , conveys that the singularities appearing at the collapse final outcome must be hidden within an event horizon and thus no distant observer could detect them . a black hole forms .
although the ccc plays a crucial role in the physics of black holes , there is yet no proof of it , due to the lack of adequate tools to treat the global characteristics of the field equations .
nevertheless , in the past 30 years many solutions to the field equations have been discovered , which exhibit the occurrence of naked singularities , where the matter content has included perfect and imperfect fluids @xcite , scalar fields @xcite , self - similar models @xcite and null strange quarks @xcite . basically , it is the geometry of trapped surfaces that decides the visibility or otherwise of the space - time singularity . in case
the collapse terminates into a naked singularity , the trapped surfaces do not emerge early enough , allowing ( otherwise hidden ) regions to be visible to the distant observers .
the gravitational collapse of scalar fields is of relevance @xcite , owing to the fact that they are able to mimic other types of behaviours , depending on the choice of the potentials .
scalar field models have been extensively examined for studying ccc in spherically symmetric models @xcite , non - spherically symmetric models @xcite and also for static cases @xcite .
their role in understanding the machinery governing the causal structure of space - time was available since the 90 s , when the numerical solutions exhibiting naked singularities were found numerically by choptuik @xcite and analytically by christodoulou @xcite .
there are in the literature a few papers discussing gravitational collapse in the presence of a scalar field joined by a fluid for the matter content@xcite : in summary , a black hole forms in these collapsing situations .
however , to our knowledge , a tachyon scalar field has not yet been considered regarding whether a black hole or naked singularity forms , that is to say , in the ccc context , together with a fluid .
tachyon fields arise in the framework of string theory @xcite and have been of recent use in cosmology @xcite .
the action for the tachyon field has a non - standard kinetic term @xcite , enabling for several effects whose dynamical consequences are different from those of a standard scalar field @xcite
. namely , other ( anti-)friction features that can alter the outcome of a collapsing scenario .
this constitutes a worthy motivation to investigate the scenario where a tachyon field is added to a barotropic fluid , both constituting the matter content present in the collapse process : on the one hand , the fluid will play the role of conventional matter from which a collapse can proceed into , whereas , on the other hand , the tachyon would convey , albeit by means of a simple framework , some intrinsic features from a string theory setting .
restricting ourselves herein to the tachyon as the intrinsic string ingredient influencing the collapse , let us nevertheless point that many other string features could be incorporated in subsequent similar studies @xcite .
our purpose , in this paper , by investigating the gravitational collapse of a barotropic fluid together with a tachyon field , is therefore to establish the types of final state that can occur ( i.e. , whether a black hole or a naked singularity emerges , in the context of the ccc ) , which matter component will determine the outcome . in particular , if the late time tachyon behaviour , possibly competing with the fluid and eventually becoming dominant , could allow interesting features to appear .
we then organize this paper as follows .
in section [ collapse ] we give a brief review on the gravitational collapse of a specific space - time , namely the marginally bounded case ( cf .
@xcite ) . in section [ classic ]
we study , by means of a dynamical system analysis , the gravitational collapse employing a tachyon and a barotropic fluid as the matter content .
the analytical study is complemented by a careful numerical investigation . in section [ discussion ]
we present our conclusions and a discussion of our results .
in this section , we will discuss the space - time region inside the collapsing sphere which will contain the chosen matter content .
an isotropic friedmann - robertson - walker ( frw ) metric , in comoving coordinates , will be considered as the interior space - time for the gravitational collapse . however , in order to study the whole space - time , we must match this interior region to a suitable exterior . in the model herein ,
it is convenient to consider a spherically symmetric and inhomogeneous space - time such as the schwarzschild or the generalized vaidya geometries to model the space - time outside the collapsing sphere . in order to further illustrate this specific gravitational collapse process ,
let us employ a very particular class of solutions .
we assume that the behavior of the matter density , for @xmath32 , is given by the following ansatz ) is spatially homogeneous , it is obvious that , the energy - momentum tensor must also be spatially homogeneous . ] ( cf . ref .
@xcite ) , @xmath33 where @xmath34 and @xmath35 are positive constants ) and pressure ( [ pressapprox ] ) , the weak energy condition can be written as follows @xmath36 the weak energy condition is satisfied throughout the collapse process ( see also next section for more details regarding the issue of the energy conditions ) . ] .
integration of the first equation in ( [ einstein-1 ] ) gives the following relation for the mass function as @xmath37 hence we have @xmath38 it is therefore possible to identify the following outcomes : * for @xmath39 , the ratio @xmath40 is less than unity throughout the collapse and if no trapped surfaces exist initially , due to the regularity condition ( @xmath41 for a suitable value of @xmath34 and @xmath5 ) , then no trapped surfaces would form until the epoch @xmath42 .
more precisely , there exists a family of radial null trajectories emerging from a naked singularity . * for @xmath43
, the ratio @xmath44 goes to infinity and trapped surfaces will form as the collapse evolves , which means that singularity will be covered and no radial geodesics can emerge from it .
thus , a black hole forms . from equation ( [ energyapprox ] ) , we may easily find a relation between the energy density and pressure as @xmath45 which shows that for @xmath46 , where a naked singularity occurs , the pressure gets negative values @xcite . from equations ( [ einstein2 ] ) , ( [ energyapprox ] ) and bearing in mind the continuous collapse process @xmath47 we can solve for the scale factor as @xmath48 where @xmath49 is the time at which the energy density begins increasing as @xmath50 and @xmath51 corresponds to @xmath49 . assuming that the collapse process initiates at @xmath52 , the time at which the scale factor vanishes corresponds to @xmath53 , implying that the collapse ends in a finite proper time .
the model we employ to study the gravitational collapse has an interior space - time geometry as described above ( see e.g. @xcite ) . concerning the matter content
, we consider a spherically symmetric homogeneous tachyon field together with a barotropic fluid .
we use an inverse square potential for the tachyon field given by @xcite , @xmath54 where @xmath55 is a constant .
note that we can consider the potential ( [ potential ] ) as two ( mirror ) branches ( see figure [ pot - v2 ] ) , upon the symmetry @xmath56 , treating the @xmath57 and @xmath58 separately .
the total energy density of the collapsing system is therefore @xmath59 , with @xmath60 where @xmath61 is the energy density of the barotropic matter , whose pressure @xmath62 , in terms of the barotropic parameter , @xmath2 , satisfies the relation @xmath63 , the barotropic parameter being positive , @xmath64 .
then , the raychadhuri equation for the collapsing system can be written as @xmath65 furthermore , the equation of motion for the tachyon field can be obtained as @xmath66.\label{field}\ ] ] the energy conservation of the barotropic matter is @xmath67 integrating ( [ consenrbaro ] ) , the energy density for barotropic matter can be presented as @xmath68 , where @xmath69 is a constant of integration .
the total pressure is given by @xmath70 in order to have a physically reasonable matter content for the collapsing cloud , the tachyon field and the barotropic fluid would have to satisfy the weak and dominant energy conditions @xcite : for any time - like vector @xmath71 , the energy - momentum tensor @xmath72 satisfies @xmath73 and @xmath74 must be non - space - like .
since for the symmetric metric ( [ metric ] ) we have @xmath75 and @xmath76 , then , for the energy densities and the pressures of the tachyon field and the barotropic fluid the energy conditions amount to the following relations @xmath77 the first two expressions in equations ( [ ec1 ] ) and ( [ ec2 ] ) denote the ` weak energy condition ' ( wec ) , and the last expression denotes the ` dominant energy condition ' ( dec ) . for a tachyon field , it is straightforward to show that the wec is satisfied : @xmath78 in addition , since @xmath79 , then @xmath80 from which it follows that @xmath81 , and hence , the dec is satisfied for the tachyon matter . for the barotropic fluid
considered herein , we assume a regular initial condition for the collapsing system , such as @xmath82 , indicating that @xmath83 .
in addition , since @xmath84 , then @xmath85 provided that @xmath64 ; through this paper we will consider the positive barotropic parameter @xmath2 .
hence , the wec is satisfied . on the other hand ,
considering the dec ( @xmath86 ) it follows that the barotropic parameter @xmath2 must satisfy the range @xmath87 ( cf . the phase space analysis in section [ classic1 ] ) .
we analyze in this ( sub)section the gravitational collapse of the above setup by employing a dynamical system perspective .
we introduce a new time variable @xmath88 ( instead of the proper time @xmath4 present in the comoving coordinate system @xmath89 ) , defined as @xmath90 where @xmath91 ( the limit @xmath92 corresponds to an initial condition of the collapsing system ( @xmath93 ) and the limit @xmath94 corresponds to @xmath95 ) . for any time dependent function @xmath96 ,
we can therefore write , i.e. , @xmath13 is assumed . ] @xmath97 we further introduce a new set of dynamical variables @xmath98 , @xmath99 and @xmath100 as follows @xcite @xmath101 from the new variables ( [ dyn1 ] ) , ( [ dyclxy ] ) and the system ( [ raychad1baro])-([consenrbaro ] ) with ( [ potential ] ) , an autonomous system of equations is retrieved : @xmath102,\label{dyclybaro}\\ \frac{ds}{d\tau } & \equiv f_{3}=s(1-s)(\gamma - x^{2}).\label{dyclsbaro}\end{aligned}\ ] ] note that @xmath103 is given by @xmath104 which for the potential ( [ potential ] ) brings @xmath105 as a constant ; for the @xmath57 branch , @xmath106 , and for @xmath58 branch , @xmath107 . equation ( [ energy ] ) in terms of the new variables reads . ] , @xmath108 the dynamical variables have the range @xmath109 , and @xmath110 .
this brings an effective two - dimensional phase space . setting @xmath111
, we can obtain the critical points @xmath112 for the autonomous system .
the stability can be subsequently discussed by using the eigenvalues of the jacobi matrix @xmath113 , defined at each fixed point @xmath112 , as @xmath114 solutions , in terms of the dynamical variables , in the neighborhood of a critical point @xmath115 can be extracted by making use of @xmath116 with the perturbation @xmath117 given by @xmath118 where @xmath119 , @xmath120 are the eigenvalues of the jacobi matrix , and the @xmath121 are constants of integration . we have summarized the fixed points for the autonomous system and their more relevant stability properties in table [ t1 ] .
.summary of critical points and their properties for the @xmath58 ( denoted with square brackets , i.e. , as , for example , @xmath122 $ ] ) and @xmath57 branches ( denoted with regular brackets , i.e. , as , for example , @xmath123 ) . [ cols="^,^,^,^,^,^",options="header " , ] table [ t1 ] is complemented with figure [ 3d - stream ] .
the autonomous system equations ( [ dyclxbaro])-([dyclsbaro ] ) unfolds a three dimensional phase space constrained by ( [ dyclfriedbaro ] ) .
therefore the main fixed points ( and the trajectories nearby ) are located on surfaces of this three dimensional phase space .
furthermore , from table [ t1 ] , we can focus our attention on three situations characterized by having @xmath124 , @xmath125 and @xmath126 .
briefly , in more detail : * * point @xmath127 * : the eigenvalues are @xmath128 , @xmath129 and @xmath130 . for @xmath131 ,
all the characteristic values are real and negative , then the trajectories in the neighborhood of this point are attracted towards it .
hence , @xmath127 is a stable node ( attractor ) .
finally , for @xmath132 , all characteristic values are real , but one is positive and two are negative , the trajectories approach this point on a surface and diverge along a curve : this is a saddle point .
+ * * point @xmath133 * : the eigenvalues are @xmath128 , @xmath129 and @xmath130 .
this point has the same eigenvalues of point @xmath127 and similar asymptotic behavior , being also a stable node for @xmath131 ( see the left plot in figure [ 2d - stream ] ) and a saddle point for @xmath132 .
* * point @xmath134 : * this fixed point has eigenvalues @xmath135 , @xmath136 and @xmath137 , which all are real and @xmath138 . for @xmath139 ,
two components are positive . however , from a numerical investigation we can assert that this corresponds to an unstable saddle ( see left plot in figure [ 2d - stream ] ) . on the other hand , for @xmath140 , one component is negative and other is positive , and hence , a saddle point configuration would emerge .
* * point @xmath141 : * the eigenvalues are @xmath142 , @xmath143 and @xmath144 . as @xmath64 , trajectories approach this point on a surface ( the in - set ) and diverge along a curve ( the out - set ) .
this is a saddle _ _ point ( see left plot in figure [ 3d - stream ] ) .
* * point @xmath145 : * this point is located at @xmath146 , where @xmath147 . from the constraint @xmath148
we get @xmath149 i.e. , @xmath150 when @xmath151 , points @xmath152 and @xmath153 become coincident ( see table [ t1 ] ) .
when we consider @xmath154 , we obtain @xmath155 , @xmath156 and @xmath157 .
consequently point @xmath152 and @xmath158 become coincident .
therefore , point @xmath152 can be found along a curve that joins points @xmath158 ( on the surface with @xmath124 ) and @xmath153 ( on the surface with @xmath125 ) .
the eigenvalues are @xmath135 , with @xmath159 and @xmath160 given by , @xmath161 the eigenvalues ( [ eigencle ] ) are non negative for @xmath140 and for @xmath155 , @xmath162 and @xmath163 .
+ + in the plot at the center of figure [ 2d - stream ] we provide a section of the phase space showing point @xmath152 , where it can be seen that all trajectories are divergent from it .
this behavior is characteristic of an unstable node and expected whenever point @xmath152 is found along the curve that connects @xmath158 and @xmath153 .
* * point @xmath164 : * the eigenvalues are @xmath128 , @xmath143 and @xmath165 . for @xmath132 ,
all the characteristic values are real and negative , then the trajectories in the neighborhood of this point are attracted towards it .
hence , @xmath164 is a stable node ( attractor ) .
finally , for @xmath131 , all characteristic values are real , but one is positive and two are negative .
the trajectories approach this point on a surface and diverge along a curve ; this is a saddle point . *
* point @xmath166 : * the eigenvalues are @xmath128 , @xmath143 and @xmath165 .
this point has the same eigenvalues of point @xmath164 and similar asymptotic behavior , being also a stable node for @xmath132 ( see the right plot in figure [ 2d - stream ] ) and a saddle point for @xmath131 . *
* point @xmath167 : * this point is found along the line segment with @xmath168 , @xmath169 and @xmath1700,1\right[$ ] .
the eigenvalues are @xmath128 , @xmath129 and @xmath171 , where @xmath1720,1\right[$ ] . in this case , all the characteristic values are real and negative , then the trajectories in the neighborhood of this points are attracted towards it .
hence , @xmath167 is a stable node ( attractor ) . * * point @xmath173 : * this point is found along the line segment with @xmath174 , @xmath169 and @xmath1700,1\right[$ ] .
the eigenvalues are @xmath128 , @xmath129 and @xmath171 .
this point has the same eigenvalues of point @xmath167 and similar asymptotic behavior , being also a stable node . before proceeding ,
let us add two comments .
on the one hand , note the transition that occurs , when going from @xmath131 to @xmath132 , an intermediate state as discussed ( by means of a numerical study ) in subsection [ tracking - solutions ] .
this bifurcation behaviour is made explicit through the numerical methods employed .
these allowed us to confirm the results on the dynamical system analysis and to further assess in regions like the transition from tachyon dominance to fluid dominance , verifying all possible scenarios with those two tools .
on the other hand , figure [ 3d - stream ] deserves some attention when we consider the trajectories approaching the stable nodes @xmath175 ( left plot ) and @xmath176 ( right plot ) . therein , the trajectories approach @xmath177 along a segment line containing the stable nodes .
this situation implies that the energy density ( [ energy ] ) diverges after reaching a point where @xmath177 , @xmath178 and @xmath179
. this observation will be important in the next sections concerning the possible outcomes of the gravitational collapse .
let us herewith discuss this section possible outcomes regarding the collapsing system , employing elements from both our analytical as well of numerical study . from the trajectories in the vicinity of @xmath127 and @xmath133
, attractor solutions can be described .
the asymptotic behavior of @xmath100 near the point @xmath127 can be approximated as @xmath180 ; hence , as @xmath94 ( i.e. , @xmath95 ) , @xmath100 vanishes .
moreover , the time derivative of the tachyon field is given by @xmath181 , that is , the tachyon field @xmath182 has a linear time dependence and can be approximated as ( see figure [ f3b ] ) @xmath183 @xmath184 it should be noted that , for the @xmath57 branch , @xmath185 is positive on the initial configuration of the collapsing system ( where @xmath186 ) and hence , the tachyon field increases with time , proceeding downhill the potential . within a finite amount of time , the tachyon field reaches its maximum . ] but finite value @xmath187 at @xmath188 , with the minimum ( but non - zero value ) @xmath189 . as the tachyon potential decreases , the dynamical variable @xmath190 vanishes .
thus , as @xmath191 , the energy density of the system diverges .
furthermore , the tachyon pressure @xmath192 vanishes asymptotically branch , @xmath185 is negative at the initial condition .
thus , the absolute value of the tachyon field starts to decrease from the initial configuration as @xmath193 until the singular point at time @xmath194 , where tachyon field reaches its minimum but non - zero value @xmath195 .
this leads it uphill the potential until the singular epoch , where the potential becomes maximum but finite . ]
( see figure [ f3 ] for plots of the energy density , barotropic pressure and tachyon field pressure ) .
the total energy density of the collapsing system is given approximately by the energy density of tachyon field ( see figure [ dominance ] for numerical solutions ) , which corresponds to a dust - like matter near the singularity as @xmath196 . from this relation for the tachyon energy ,
we may write the energy density of the system near point @xmath127 , @xmath197 this induces a black hole at the collapse final state .
as far as @xmath198 is concerned , the time dependence of the tachyon field can be obtained as @xmath199 for the @xmath57 branch , the tachyon field decreases from its initial value at @xmath186 ( where @xmath200 ) , moving uphill the potential .
then , the potential will reach a maximum ( but finite ) value when the tachyon field reaches its minimum and nonzero value ( i.e. @xmath201 as @xmath202 ) . in the limit case @xmath203 ,
the tachyon field vanishes and the potential diverges .
i.e. , @xmath99 and @xmath100 vanish , the hubble rate increases faster than the potential and the barotropic fluid energy density diverges , that is , the total energy density of the system asymptotically diverges .
branch , the tachyon field increases from its initial condition as time evolves , proceeding to ever less negative values as @xmath204 . in this case , it proceeds downhill the tachyon potential till the system reaches @xmath205 at @xmath206 , where the hubble rate and hence the total energy density of the system diverge .
this implies that the time at which the collapse system reaches the singularity is always @xmath207 . ] .
likewise , when @xmath191 and @xmath95 , the tachyon matter behaves as dust matter .
the fate of the collapse for this fixed point is as well a black hole formation .
the asymptotic solutions provided by the fixed points @xmath127 and @xmath133 correspond to a dust - like solution with a vanishing pressure for the tachyon field , whose energy density reads @xmath196 .
this is consistent with the wec and dec for the tachyon matter being satisfied , as was mentioned before .
concerning the status of the energy conditions for the barotropic fluid , as we indicated before , regularity of the initial data for the collapsing matter respects the wec . on the other hand , stability of the solution in this case
ensures that @xmath131 , which satisfies the sufficient condition for the dec . from the trajectories in the vicinity of @xmath164 and @xmath166
, solutions can be described with some having an attractor behaviour .
the asymptotic behavior of @xmath100 near the point @xmath164 can be approximated as @xmath208 ; hence , as @xmath94 ( i.e. , @xmath95 ) , @xmath209 .
moreover , the time derivative of the tachyon field is given by @xmath181 ; that is , the tachyon field @xmath182 has a linear time dependence and can be approximated as ( see figure [ f3b ] ) , @xmath183 the total energy density of the collapsing system is given approximately by the energy density of the barotropic fluid ( see figure [ dominance ] , where @xmath210 while @xmath211 for @xmath132 ) , which goes , near the singularity , as @xmath212 . from this relation for the fluid energy ,
we may write the energy density of the system near point @xmath164 , @xmath213 for @xmath214 , the ratio @xmath215 , converges as the singularity is reached leading to the avoidance of trapped surfaces , but since the corresponding fixed point @xmath164 turns to be a saddle , then the resulting naked singularity is not stable .
for @xmath216 the ratio @xmath44 diverges and the trapped surfaces do form .
but still the point @xmath164 is saddle and the resulting black hole is not stable .
the case @xmath132 corresponds to a stable solution for which the ratio @xmath44 goes to infinity as the collapse advances .
then , the trapped surface formation in the collapse takes place before the singularity formation and thus the final outcome is a black hole .
as far as @xmath217 is concerned , the time dependence of the tachyon field can be obtained as @xmath199 for the @xmath57 branch , the tachyon field decreases from its initial value at @xmath186 ( where @xmath200 ) , moving uphill the potential , the potential will reach a maximum ( but finite ) value when the tachyon field reaches its minimum and nonzero value ( i.e. @xmath201 as @xmath202 ) . in the limit case @xmath203 ,
the tachyon field vanishes and the potential diverges .
i.e. , @xmath99 and @xmath100 vanish , the hubble rate increases faster than the potential and the barotropic fluid energy density diverges , that is , the total energy density of the system , is given by equation ( [ tenerg - a-1 ] ) , and asymptotically diverges .
similar to point @xmath164 the mass function of the system for the fixed point solution @xmath166 is given by @xmath215 , which diverges for @xmath132 .
therefore , the resulting singularity in this case will be covered by a black hole horizon . as far as the energy conditions are concerned for the fixed point solutions @xmath164 and @xmath166
, we find that the tachyon field satisfies the wec .
also the dec remains valid case as well . on the other hand , for the barotropic fluid , the wec is satisfied initially and will hold until the endstate of the collapse .
the stable solution in this case corresponds to the range @xmath132 which satisfies dec as well . in this subsection
, we will discuss a different type of solutions , where the fluid and tachyon appear with a tracking behaviour .
let us introduce this situation as follows .
interesting and physically reasonable tracking solutions can be found , where @xmath218 , when we consider @xmath219 , i.e. , a situation whereby the emergence of points @xmath220 and @xmath221 will be of relevance as attractors . the transition from tachyon dominated to fluid dominated scenarios , like those described in previous sections , is not straightforward . in this situation ,
the tachyon field and barotropic fluid compete to establish the dominance in the late stage of the collapse . in figure [ dominance ] we have a illustration of this kind of solutions .
the mentioned dominance seems to depend strongly on the initial ratio @xmath222 at an earlier stage of the collapse and also on the value of @xmath2 .
such a dependence on the initial conditions can lead to a set of solutions between those provided by fixed points @xmath123 , @xmath198 ( tachyon dominated solutions ) and by points @xmath223 , @xmath217 ( fluid dominated solutions ) . from a dynamical system point of view , this corresponds to have trajectories asymptotically approaching @xmath177 in sections where @xmath100 is between 0 and 1 ( see figure [ 3d - stream ] ) . at the end of the collapse
we observe that @xmath224 , as illustrated in figure [ dominance ] .
therefore , in this scenario , the trajectories would convey a collapsing case in which the energy density of the tachyon field and of the barotropic fluid are given by @xmath225 .
this shows a tracking behavior for the collapsing system @xcite .
moreover , the total energy density of the collapse , in terms of @xmath226 , reads @xmath227 equation ( [ energ - e ] ) shows that the energy densities of the tachyon field , the barotropic matter and hence , their total ( for the collapsing system ) , diverges as @xmath95 .
the ratio of the total mass function over the area radius is given by @xmath228 equation ( [ massd ] ) subsequently implies that , for an adequate choice of values @xmath229 , trapped surfaces can form as the collapse evolves and a few scenarios can be extracted .
more precisely , for the range @xmath230 , for both @xmath57 and @xmath58 branches , the final fate of the collapse is a black hole . for the case
in which @xmath231 , the ratio @xmath44 remains finite as the collapse proceeds and an apparent horizon is delayed or fails to form ; the final state is a naked singularity ( a solution for the choice of ` @xmath232 ' sign in equation ( [ eigencle ] ) ) .
the tracking solution indicates @xmath233 as the threshold ( illustrated in figure [ 2d - stream ] ) , which distinguishes a black hole or a naked singularity forming .
therefore , under suitable conditions , we can determine whether it is possible to have the formation of a naked singularity .
in fact , if we assume a very unbalanced initial ratio @xmath222 with @xmath234 and a barotropic fluid having @xmath231 , then we can have a situation where the ratio ( [ massd ] ) is converging .
the set of initial conditions described by @xmath234 are equivalent to consider the barotropic fluid as initially dominant .
if this specific unbalanced distribution of matter is allowed to evolve into a regime where the tachyon dominates then the system will evolve until @xmath235 becomes comparable to @xmath236 .
the singularity is reached in finite time and it can happen before the tachyon can effectively dominate .
in figure [ naked singularity ] we have a graphical representation of the ratio @xmath44 .
it can be seen that the ratio @xmath44 remains finite for @xmath214 , while the energy density is diverging , as the collapse proceeds and apparent horizon is delayed or fails to form till the singularity formation .
as the right plot shows , the validity of wec is guaranteed throughout the collapse scenario for both barotropic fluid and tachyon field .
also the dec is valid for the solutions that exhibit naked singularity , i.e. , those for which @xmath214 .
[ naked singularity ] point @xmath134 does not represent an attractor and hence its vicinity will not corresponds to a collapse endstate .
however , trajectories emerging from it would proceed towards @xmath127 or @xmath133 .
the time dependence of the tachyon field near @xmath134 can be obtained integrating @xmath237 as @xmath238 in the @xmath57 branch , the tachyon field starts to increase with time from its initial value at @xmath186 ( where @xmath239 ) , proceeding uphill the potential .
then , a stage is obtained where a maximum ( but finite ) value is reached , when the tachyon field goes to its minimum and nonzero value at @xmath240 .
asymptotically , @xmath241 and @xmath242 .
since @xmath243 is a constant , the hubble rate at @xmath206 reads @xmath244 and the total energy density of the system is essentially the energy density of the tachyon field and is given by @xmath245 which is proportional to the tachyon potential .
therefore , at the time @xmath246 , when the tachyon field reaches its non - zero minimum , the total energy density of the system remains finite . in the limit case , when the time @xmath247 , the tachyon field vanishes , the tachyon potential diverges .
alternatively , we can also write that tachyon behaves as @xmath248 when @xmath95 , the tachyon field vanishes , and hence , the potential and energy density of tachyon field in equation ( [ cenerg ] ) , as a function of @xmath226 , is given by @xmath249 and diverges at @xmath205 .
the dynamics of the system in vicinity of this fixed point is given by @xmath250 , @xmath251 and @xmath252 , where @xmath162 .
concerning @xmath141 , the energy density of the tachyon field is approximately given by the tachyon potential @xmath253 ( with @xmath254 ) and thus , in its vicinity , the tachyon potential remains constant and finite .
furthermore , @xmath255 and @xmath256 imply that the total energy density is given by the energy density of barotropic matter only : @xmath257 .
this solution diverges as @xmath95 .
the fixed point @xmath145 for the range @xmath140 presents solutions for the collapsing system in which @xmath258 asymptotically is given by @xmath259 , for the @xmath57 and @xmath260 for the @xmath58 branch . for the @xmath57 branch ,
the time dependence of the tachyon field is given by @xmath261 the tachyon starts its evolution from the initial configuration @xmath200 , and decreases uphill the potential .
then , at the time @xmath262 the tachyon field vanishes , the potential diverging branch , the time dependence of the tachyon field is given by @xmath263 .
therefore the tachyon field starts its evolution from the initial configuration at @xmath264 , decreasing in time , going uphill the potential . ] .
a thorough numerical analysis of the trajectories in the vicinity of point @xmath152 has shown that it behaves as an unstable node .
we can observe in the center plot of figure [ 2d - stream ] an illustration of the general behavior near this point , namely having trajectories diverging from it . in order to complete the full space - time geometry for the herein collapsing model
, we need to match the homogeneous interior space - time to a suitable ( inhomogeneous ) exterior geometry . for
a perfect fluid gravitational collapse set up , with equation of state @xmath265 , the pressure does not necessarily vanish at the boundary .
e.g. , matching the internal geometry filled with matter ( and radiation ) to a boundary layer ( which is crossed with the radiation ) , which could in turn be matched to an exterior geometry , not completely empty ( e.g. , filled by radiation ) . more concretely , matching the interior with a generalized exterior vaidya space - time across the boundary given by @xmath266 @xcite . similarly to @xcite ,
we proceed by considering the interior metric which describes the collapsing cloud as @xmath267 and the exterior one in advanced null coordinates @xmath268 given by @xmath269 in these coordinates the outermost boundary of trapped surfaces is simply given by the relation @xmath270 .
it should be noticed that in general the formation of the singularity at @xmath205 is independent of matching the interior to the exterior space - time . in order to find a suitable exterior metric function @xmath271 we resort to an hamiltonian perspective of the model .
the total hamiltonian constraint is given by @xcite @xmath272 where @xmath273 and @xmath274 are the conjugate momentums for the scale factor @xmath226 and for the tachyon field @xmath0 , respectively .
furthermore , the hamilton equations for parameters @xmath226 , @xmath273 and @xmath274 can be obtained by using the equation ( [ hamiltonian - cl ] ) as follows : @xmath275 on the other hand , since the hamiltonian constraint , @xmath276 , must be held across the boundary @xmath277 , equation ( [ hamiltonian - cl ] ) reduces to @xmath278 whereby , substituting for @xmath273 and @xmath274 from equation ( [ eqs-1])-([eqs-3 ] ) , we get @xmath279 furthermore , it is required for the junction condition @xmath280 and the equation of motion for scale factor @xmath281 to be satisfied at the boundary of two regions . by substituting these conditions in the hamiltonian constraint equation , @xmath282 , given by ( [ fex ] )
, we get @xmath283=\left(\frac{r}{r_{b}}\right)^{4}\left[\frac{v^{2}}{1-\dot{\phi}^{2}}-\rho_{0\text{b}}^{2}\left(\frac{r}{r_{b}}\right)^{-6\gamma}\right],\label{area}\end{aligned}\ ] ] notice that , we have substituted the four - velocity of the boundary being seen from the exterior by @xmath284 ( cf .
then , by solving equation ( [ area ] ) for @xmath285 , the boundary function is obtained simply as @xmath286 let us now study the behavior of the boundary function for the stable fixed point solutions we obtained in the previous subsections .
we note that the above expression is valid when both the tachyon field and barotropic fluid are present , thus in the regimes where the tachyon field is dominant the contribution of the fluid to hamiltonian constraint is set aside . for the dust - like solutions , described by points @xmath123 and @xmath198 , the energy density of tachyon field at the boundary @xmath287 , included in second term of equation ( [ exteriormf ] ) , is given by @xmath288 .
so , the boundary function @xmath285 in equation ( [ exteriormf ] ) reduces to : @xmath289 and @xmath290 .
equation ( [ exteriormf - dust ] ) for @xmath291 constrains the exterior space - time to have a schwarzschild metric in advanced null coordinates , providing an interpretation of the collapsing system as a dust ball with the radius @xmath287 and the density @xmath292 . on the other hand , for the fixed point solution @xmath145 , by using the equation ( [ energ - e ] ) in ( [ exteriormf ] ) , we have @xmath293 where @xmath294
( notice that , this corresponds to the case in which @xmath295 in previous subsection ) . in order to get a possible class of dynamical exterior solutions , we proceed by considering the following metric at the boundary @xcite @xmath296 where @xmath297 is the total mass within the collapsing cloud . matching the for exterior metric function gives @xmath298 in which , the second part stands for matching for the extrinsic curvature .
differentiation of the first expression in ( [ matching ] ) with respect to time and using the second one , we get @xmath299 multiplying the first expression in equation ( [ matching ] ) by @xmath300 and after adding the result with the above equation , we get @xmath301 now , by substituting for @xmath302 into equation ( [ matching ] ) , we obtain @xmath303 then , as seen from equation ( [ massd ] ) for @xmath214 , where the trapping of light has failed to occur , the exterior geometry is dynamical in contrast to the tachyon dominated regime in which a space - like singularity forms with a static exterior space - time . finally , for fluid dominated solutions , depending on the value of @xmath2 , both naked singularities and black holes may form , the mass being different to those of tracking solutions and the exterior geometry being static or dynamical , respectively .
in this paper we considered a particular setting among the spherically symmetric class of models for gravitational collapse , with a tachyon field and a barotropic fluid as matter content .
we restricted ourselves to the marginally bound case ( cf .
@xcite for a description and details ) .
the tachyon potential was assumed to be of an inverse square form .
our objective was to establish ( i ) which final states would occur ( namely , a black hole or a naked singularity ) , ( ii ) how each matter component will compete ( the fluid being conventional , whereas the tachyon bringing some workable but intrinsic non - standard effects from string theory ) and , ( iii ) which will eventually be the determinant component at the end of the collapse process .
more precisely , in our opinion it is of interest to assert , if and how , at later stages , effects induced by the tachyon ( a scalar field ( among others ) found in string theory context ) , could allow interesting features to be eventually discussed ( i.e. , what asymptotic behaviour emerges ) .
as far as the tachyon field is concerned , from eq .
( [ field ] ) we can have that with @xmath304 and @xmath305 , @xmath258 terms act like an anti - friction contribution at a collapsing phase , within an uphill evolution for the @xmath306 potential for tachyon field ( i.e. , when @xmath307 ) . moreover , other terms ( i.e. the term including @xmath308 ) would have an anti - friction effect as well . determining therefore the outcome of the gravitational collapse in our system , i.e. ,
whether , e.g. , a black hole or naked singularity would form , was not a straightforward assessment .
we considered an analytical description by means of a phase space analysis @xcite , discussing several asymptotic behaviours ; these were also subject to a careful study involving a numerical investigation , which added a clearer description of the possible dynamical evolutions . within this setting ,
for a spatially homogeneous interior space - time , we found a situation where the tachyon was dominant , with a black hole forming i.e. , @xmath309 . in both models , the collapsing systems lead to the black hole formation . ] .
a cosmological framework involving only a frw geometry , with a tachyon field and a barotropic fluid , was investigated in @xcite , focusing on the late time ( dark energy like ) stages .
it is interesting to note that while for a tachyon dominated regime , an inflationary - like ( i.e. , accelerated expansion ) scenario leads to violation of the strong energy condition ( sec ) .
thus , the sec holds if @xmath310 . ]
@xcite , in our collapse scenario we have instead a corresponding asymptotic stage , where tachyon dominance leads to black hole formation , satisfying the sec at the final state of the collapse ( cf .
figure ( [ f3b ] ) ) .
the same behaviour is observed in the fluid dominated regime , where for @xmath132 the sec is satisfied by the fluid , with a black hole formation .
however , for the tracking solutions , where a naked singularity forms , i.e. , @xmath311 and @xmath214 , the sec is violated . moreover , in further comparison with the set of fixed points found in @xcite for an expanding ( accelerating ) universe ( @xmath312 ) , we have found analytically two additional critical points ( @xmath285 and @xmath313 ) for the collapse process which correspond to a barotropic dominated collapsing regime which ends in a black hole for @xmath132 , respecting the wec and sec .
indeed , for the late - time acceleration of the universe , filled with a tachyon field and a fluid , the phase space analysis in @xcite predicted two class of solutions : a tachyon dominated solution ; and a tracking solution @xcite .
rather differently , corresponding solutions herein our paper turn to be unstable towards the singularity in the collapse process .
nevertheless , two other solutions are also provided in our collapsing scenario : a tachyon dominated solution where @xmath2 satisfies the range @xmath131 , initially ; and a barotropic dominated solution for which the barotropic parameter holds the range @xmath132 .
all these solutions predict a black hole formation as collapse end state . further regarding the tracking solutions ( of a cosmological nature ) indicated in @xcite . in our collapsing system ,
a different and rather interesting set of states is found , within the context of tracking behavior for the barotropic fluid plus tachyon field content .
being more concrete , these solutions have that a black hole or naked singularity forms .
in particular , in this situation , we found that it is possible to define by a numerical appraisal in particular , the threshold @xmath314 , separating black hole and naked singularity solutions for the gravitational collapse .
we have also discussed the specific conditions leading to the formation of a naked singularity .
therefore , we concluded , that in our model , if the collapse starts with an unbalanced distribution of the matter content favoring the barotropic fluid , i.e. @xmath315 , then towards the final stage of the collapse they evolve until @xmath316 , when a naked singularity forms .
however , since the nec is satisfied , the strong curvature condition along the null geodesics can be preserved and the singularity can be strong in the sense of @xcite .
we think it is fair to indicate that we employed a @xmath306 potential for the tachyon , whereas for @xmath317 , the tachyon should not induce a divergent behavior as far as string theory advises @xcite .
in fact , an exponential - like potential for the tachyon could bring a richer set of possible outcomes @xcite , including a better behaved and possibly a more realistic evolution when dealing with @xmath317 .
finally , let us add that it will be of interest to investigate ( i ) other scenarios for the geometry of the interior space - time region , ( ii ) specific couplings between the tachyon and the fluid , within e.g. , a chamaleonic scenario for gravitational collapse and black hole production ( broadening the scope in @xcite ) , ( iii ) adding either axionic , dilatonic or other terms ( e.g. , curvature invariants ) that could be considered from a string setup , but at the price of making the framework severely less workable and ( iv ) whether explicit quantum effects can alter the outcomes presented in this paper . to this purpose latter
, we plan to use ingredients brought from loop quantum gravity ( cf . ref .
the authors would like to thank p. joshi , a. khaleghi , c. kiefer , f. c. mena , a. vikman , and j. ward for making useful suggestions and comments .
they are also grateful to the referee for the useful comments and suggestions on the issue of energy conditions .
yt is supported by the portuguese agency fundao para a cincia e tecnologia through sfrh / bd/43709/2008 .
this research work was also supported by the grant cern / fp/109351/2009 and cern / fp/116373/2010 .
t. harada , phys .
d * 58 * ( 1998 ) 104015 ; + t. harada and h. maeda , phys .
d * 63 * ( 2001 ) 084022 ; + r. goswami and p. s. joshi , class .
quantum grav * 19 * ( 2002 ) 5229 ; + r. giamb , f. giannoni , g. magli and p. piccione , gen .
grav * 36 * ( 2004 ) 1279 ; + j. f. villas da rocha and a. wang , class .
quantum grav * 17 * ( 2000 ) 2589 ; + r. giamb , f. giannoni , g. magli and p. piccione , class .
quantum grav * 20 * ( 2003 ) 4943 .
p. szekeres and v. iyer , phys .
d * 47 * ( 1993 ) 4362 ; + s. barve , t. p. singh , and l. witten , gen .
* 32 * ( 2000 ) 697 ; + a. a. coley and b. o. j. tupper , phys .
d * 29 * 2701 ( 1984 ) ; + k. lake , phys .
d * 26 * ( 1982 ) 518 .
a. i. janis , e. t. newman and j. winicour , phys .
lett * 20 * ( 1968 ) 878 ; + m. wyman , phys .
d * 24 * ( 1981 ) 839 ; + b. c. xanthop oulos and t. zannias , phys .
d * 40 * ( 1989 ) 2564 ; + o. bergmann and r. leipnik , phys .
rev . * 107 * 1157 ( 1957 ) ; + h. a. buchdahl , phys .
rev * 115 * ( 1959 ) 1325 . p. s. joshi and r. goswami , class .
quantum grav .
* 24 * 2917 ( 2007 ) ; + r. goswami , p. s. joshi , c. vaz and l. witten , phys.rev .
d * 70 * 084038 ( 2004 ) ; + p.
s. joshi and r. goswami , arxiv:0711.0426v1 [ gr - qc ] h. k. khalil , _ nonlinear systems _ ,
2nd edition ( englewood cliffs .
nj : prentice hall , 1996 ) ; jack carr , _ applications of center manifold theorem _
( springer - verlag , 1981 ) ; j. guckenheimer , p. holmes , _ nonlinear oscillations , dynamical systems and bifurcation of vector fields _ ( springer - verlag , 1983 ) . f. j. tipler , phys . lett . a 67 ( 1977 ) 8 ; f. j. tipler , c. j. s. clarke , and g. f. r. ellis , in general relativity and gravitation , edited by a. held ( plenum , new york , 1980 ) , vol .
2 , p. 97 ; c. j. s. clarke , the analysis of space - time singularities ( cambridge university press , cambridge , england , 1993 ) .
a. a. sen , j. high energy phys .
* 04 * , 048 ( 2002 ) ; a. a. sen , j. high energy phys . * 07 * , 065 ( 2002 ) ; + * 16 * , 281 ( 1989 ) ; + s. h. ghate , r. v. saraykar , and k. d. patil , pramana , j. phys .
* 53 * , 253 ( 1999 ) . | a particular class of space - time , with a tachyon field , @xmath0 , and a barotropic fluid constituting the matter content , is considered herein as a model for gravitational collapse . for simplicity
, the tachyon potential is assumed to be of inverse square form _
i.e. _ , @xmath1 .
our purpose , by making use of the specific kinematical features of the tachyon , which are rather different from a standard scalar field , is to establish the several types of asymptotic behavior that our matter content induces . employing a dynamical system analysis , complemented by a thorough numerical study
, we find classical solutions corresponding to a naked singularity or a black hole formation .
in particular , there is a subset where the fluid and tachyon participate in an interesting tracking behaviour , depending sensitively on the initial conditions for the energy densities of the tachyon field and barotropic fluid .
two other classes of solutions are present , corresponding respectively , to either a tachyon or a barotropic fluid regime .
which of these emerges as dominant , will depend on the choice of the barotropic parameter , @xmath2 .
furthermore , these collapsing scenarios both have as final state the formation of a black hole . |
the interplay of stars and their surrounding gas leads to the presence of distinct phases in the interstellar medium ( ism ) of galaxies ( e.g. @xcite ) . diffuse atomic clouds ( the cold neutral medium , cnm ) have densities of about @xmath0 and temperatures of about @xmath1 , where atomic hydrogen is largely neutral but carbon is singly ionized by photons with energies between @xmath2 and @xmath3 .
the warmer ( @xmath4 ) and more tenuous ( @xmath5 ) intercloud phase is heated and ionized by fuv and euv photons escaping from hii regions @xcite , usually referred to as the warm neutral medium ( wnm ) and warm ionized medium ( wim ) .
the phases of the ism are often globally considered to be in thermal equilibrium and in pressure balance @xcite .
however , the observed large turbulent width and presence of gas at thermally unstable , intermediate temperatures attests to the importance of heating by kinetic energy input .
in addition , the ism also hosts molecular clouds , where hydrogen is in the form of @xmath6 and self - gravity plays an important role .
all of these phases are directly tied to key questions on the origin and evolution of the ism , including the energetics of the cnm , wnm and the wim ; the evolutionary relationship of atomic and molecular gas ; the relationship of these ism phases with newly formed stars ; and the conversion of their radiative and kinetic power into thermal and turbulent energy of the ism ( e.g. @xcite ) . the neutral phases of the ism have been studied using optical and uv observations of atomic lines .
these observations can provide the physical conditions but are limited to pinpoint experiments towards bright background sources and are hampered by dust extinction @xcite . at radio wavelengths , dust extinction is not important and observations of the 21 cm hyperfine transition of neutral atomic hydrogen have been used to study the neutral phases ( e.g. @xcite ) . on a global scale , these observations have revealed the prevalence of the two phase structure in the interstellar medium of cold clouds embedded in a warm intercloud medium but they have also pointed out challenges to this theoretical view @xcite .
it has been notoriously challenging to determine the physical characteristics ( density , temperature ) of the neutral structures in the ism as separating the cold and warm components is challenging ( e.g. @xcite ) . in this context ,
carbon radio recombination lines ( crrls ) provide a promising tracer of the neutral phases of the ism ( e.g. @xcite ) .
carbon has a lower ionization potential ( 11.2 ev ) than hydrogen ( 13.6 ev ) and can be ionized by radiation fields in regions where hydrogen is largely neutral .
recombination of carbon ions with electrons to high rydberg states will lead to crrls in the sub - millimeter to decameter wavelength range .
carbon radio recombination lines have been observed in the interstellar medium of our galaxy towards two types of clouds : diffuse clouds ( e.g. : @xcite ) and photodissociation regions ( pdrs ) , the boundaries of hii regions and their parent molecular clouds ( e.g. : @xcite ) . the first low frequency ( 26.1 mhz ) carbon radio recombination line was detected in absorption towards the supernova remnant cas a by @xcite ( wrongly attributed to a hyperfine structure line of @xmath7 , @xcite ) .
this line corresponds to a transition occurring at high quantum levels ( @xmath8 ) .
recently , @xcite detected crrls in the range 25.526.5 mhz towards cas a , corresponding to transitions involving levels as large as @xmath9 .
observations of low frequency carbon recombination lines can be used to probe the physical properties of the diffuse interstellar medium .
however , detailed modeling is required to interpret the observations .
@xcite showed that , at low temperatures ( @xmath10 ) , electrons can recombine with carbon ions by simultaneously exciting the @xmath11 fine structure line , a process known as dielectronic recombination - state accompanied by simultaneous excitation of the @xmath12 core electron to the excited @xmath13 state .
the captured electron can either auto ionize , collisional transferred to another state , or radiatively decay .
strictly speaking , dielectronic recombination refers to dieclectronic capture followed by stabilization .
however , throughout this article we will use the term dielectronic recombination to refer to the same process as is common in the astronomical literature . ] .
such recombination process occurs to high @xmath14 states , and can explain the behavior of the high @xmath14 crrls observed towards cas a. @xcite modified the code from @xcite to include dielectronic recombination .
@xcite modified the code to consider transitions up to 10000 levels .
all of these results assume a statistical distribution of the angular momentum levels , an assumption that is not valid at intermediate levels for low temperatures .
moreover , the lower the temperature , the higher the @xmath14-level for which that assumption is not valid . the increased sensitivity , spatial resolution , and bandwidth of the low frequency array ( lofar , @xcite ) is opening the low frequency sky to systematic studies of high quantum number radio recombination lines .
the recent detection of high level carbon radio recombination lines using lofar towards the line of sight of cas a @xcite , cyg a @xcite , and the first extragalactic detection in the starburst galaxy m82 @xcite illustrate the potential of lofar for such studies .
moreover , pilot studies have demonstrated that surveys of low frequency radio recombination lines of the galactic plane are within reach , providing a new and powerful probe of the diffuse interstellar medium .
these new observations have motivated us to reassess some of the approximations made by previous works and to expand the range of applicability of recombination line theory in terms of physical parameters .
in addition , increased computer power allows us to solve the level population problem considering a much larger number of levels than ever before .
furthermore , updated collisional rates are now available @xcite , allowing us to explicitly consider the level population of quantum angular momentum sub - levels to high principal quantum number levels .
finally , it can be expected that the square kilometer array , ska , will further revolutionize our understanding of the low frequency universe with even higher sensitivity and angular resolution @xcite . in this work ,
we present the method to calculate the level population of recombining ions and provide some exemplary results applicable to low temperature diffuse clouds in the ism . in an accompanying article ( @xcite , from here on paper ii ) , we will present results specifically geared towards radio recombination line studies of the diffuse interstellar medium . in section [ section_method ] , we introduce the problem of level populations of atoms and the methods to solve this problem for hydrogen and hydrogenic carbon atoms .
we also present the rates used in this work to solve the level population problem . in section [ section_results ]
, we discuss our results focusing on hydrogen and carbon atoms .
we compare our results in terms of the departure coefficients with previous results from the literature . in section [ section_conclusions ] , we summarize our results and provide the conclusions of the present work .
a large fraction of our understanding of the physical processes in the universe comes from observations of atomic lines in astrophysical plasmas . in order to interpret the observations , accurate models for the level population of atoms
are needed as the strength ( or depth ) of an emission ( absorption ) line depends on the level populations of atoms . here ,
we summarize the basic ingredients needed to build level population models and provide a basic description of the level population problem .
we begin our discussion by describing the line emission and absorption coefficients ( @xmath15 and @xmath16 , respectively ) , which are given by @xcite : @xmath17 where @xmath18 is the planck constant , @xmath19 is the level population of a given upper level ( @xmath20 ) and @xmath21 is the level population of the lower level ( @xmath14 ) ; @xmath22 is the line profile , @xmath23 is the frequency of the transition and @xmath24 , @xmath25 are the einstein coefficients for spontaneous and stimulated emission ( absorption ) , respectively . under local thermodynamic equilibrium ( lte ) conditions ,
level populations are given by the saha - boltzmann equation ( e.g. @xcite ) : @xmath26 where @xmath27 is the electron temperature , @xmath28 is the electron density in the nebula , @xmath29 is the ion density , @xmath30 is the electron mass , @xmath31 is the boltzmann constant , @xmath18 is the planck constant , @xmath32 is the speed of light and @xmath33 is the rydberg constant ; @xmath34 is the statistical weight of the level @xmath14 and angular quantum momentum level @xmath35 [ @xmath36 , for hydrogen ] , and @xmath37 is the statistical weight of the parent ion .
the factor @xmath38 is the thermal de broglie wavelength , @xmath39 , of the free electron .. ] in the most general case , lines are formed under non - lte conditions and the level population equation must be solved in order to properly model the line properties as a function of quantum level ( @xmath14 ) .
following e.g. @xcite and @xcite , we present the results of our modeling in terms of the departure coefficients ( @xmath40 ) , defined by : @xmath41 and @xmath42 values are computed by taking the weighted sum of the @xmath40 values : @xmath43 note that , at a given @xmath14 , the @xmath40 values for large @xmath35 levels influence the final @xmath42 value the most due to the statistical weight factor . at low frequencies stimulated emission is important @xcite and we introduce the correction factor for stimulated emission as defined by @xcite : @xmath44 unless otherwise stated the @xmath45 presented here correspond to @xmath46 transitions ( @xmath47 ) .
the description of the level population in terms of departure coefficients is convenient as it reduces the level population problem to a more easily handled problem as we will show in section [ section_levelpop ] .
the observations of high @xmath14 carbon recombination lines in the ism motivated @xcite to study the effect on the level population of dielectronic recombination and its inverse process ( autoionization ) in low temperature ( @xmath48 ) gas .
@xcite used @xmath35-changing collision rates - changing collision rates to refer to collisions rate that induce a transition from state @xmath49 to @xmath50 from @xcite and concluded that for levels @xmath51 , dielectronic recombination of carbon ions can be of importance . in a later work
, @xcite used collision rates from @xcite and estimated a value for which autoionization becomes more important than angular momentum changing rates .
the change in collision rates led them to conclude that the influence of dielectronic recombination on the @xmath42 values is important at levels @xmath52 .
clearly , the results are sensitive to the choice of the angular momentum changing rates . here
, we will explicitly consider @xmath35-sublevels when solving the level population equation .
the dielectronic recombination and autoionization processes affect only the c@xmath53 ions in the @xmath13 state , therefore we treat the level population for the two ion cores in the @xmath12 states separately in the evaluation of the level population @xcite .
the equations for carbon atoms recombining to the @xmath13 ion core population have to include terms describing dielectronic recombination ( @xmath54 ) and autoionization ( @xmath55 ) , viz .
: @xmath56 & = & \nonumber \\ \sum\limits_{n ' > n}\sum_{l'=l\pm1}{b_{n'l ' } \frac{\omega_{n'l'}}{\omega_{nl } } e^{\delta \chi_{n'n } } a_{n'l'nl}}+\sum\limits_{n ' \neq n}\sum_{l'=l\pm1}{b_{n'l ' } \frac{\omega_{n'l'}}{\omega_{nl } } e^{\delta \chi_{n'n}}(b_{n'l'nl}i_\nu + c_{n'l'nl})}+\nonumber\\ + \sum_{l'=l\pm1 } b_{nl ' } \left(\frac{\omega_{nl'}}{\omega_{nl } } \right)c_{nl'nl}+\frac{n_e n_{3/2}^+}{n_{nl}(lte)}(\alpha_{nl}+ c_{i , nl})+ \frac{n_e n_{1/2}^+}{n_{nl}(lte)}\alpha_{nl}^d.\end{aligned}\ ] ] the left hand side of equation [ eqn_llevelpopc ] describes all the processes that take an electron out of the @xmath49-level , and the right hand side the processes that add an electron to the @xmath49 level ; @xmath57 is the coefficient for spontaneous emission , @xmath58 is the coefficient for stimulated emission or absorption induced by a radiation field @xmath59 ; @xmath60 is the coefficient for energy changing collisions ( i.e. transitions with @xmath61 ) , @xmath62 is the coefficient for @xmath35-changing collisions ; @xmath63 ( @xmath64 ) is the coefficient for collisional ionization ( 3-body recombination ) and @xmath65 is the coefficient for radiative recombination .
a description of the coefficients entering in equation [ eqn_llevelpopc ] is given in section [ section_rates ] and in further detail in the appendix .
the level population equation is solved by finding the values for the departure coefficients .
the level population for carbon ions recombining to the @xmath12 level is hydrogenic and we solve for the departure coefficients ( @xmath66 ) using equation [ eqn_llevelpopc ] , but ignoring the coefficients for dielectronic recombination and autoionization .
after computing the @xmath66 and @xmath67 , we compute the departure coefficients ( @xmath68 and @xmath69 ) for both parent ion populations by summing over all @xmath35-states ( equation [ eqn_bn2 ] ) .
the final departure coefficients for carbon are obtained by computing the weighted average of both ion cores : @xmath70}{1+\left[n^+_{3/2}/n^+_{1/2}\right]}.\end{aligned}\ ] ] note that , in order to obtain the final departure coefficients , the relative population of the parent ion cores is needed . here
, we assume that the population ratio of the two ion cores @xmath71 to @xmath72 is determined by collisions with electrons and hydrogen atoms .
this ratio can be obtained using @xcite : @xmath73 where @xmath74 is the de - excitation rate due to collisions with electrons , @xmath75 is the de - excitation rate due to collisions with hydrogen atoms @xcite , @xmath76 is the atomic hydrogen density and @xmath77 is the spontaneous radiative decay rate of the core . in this work ,
we have ignored collisions with molecular hydrogen , which should be included for high density pdrs .
collisional rates for h@xmath78 excitation of c@xmath53 have been calculated by @xcite . in the cases of interest here ,
the value of @xmath79 is dominated by collisions with atomic hydrogen .
we recognize that the definition of @xmath79 given in equation [ eqpayner ] is related to the critical density ( @xmath80 ) of a two level system by @xmath81 where @xmath82 is the density of the collisional partner ( electron or hydrogen ) .
the lte ratio of the ion core is given by the statistical weights of the levels and the temperature ( @xmath27 ) of the gas : @xmath83 where @xmath84 are the statistical weights of the fine structure levels and @xmath85 is the energy difference of the fine structure transition .
the lte level population ratio as a function of temperature is shown in figure [ fig_rvalue ] , illustrating the strong dependence on temperature of this value . at densities below the critical density ( @xmath86 for collisions with h ) ,
the fine structure levels fall out of lte and the value for @xmath79 becomes very small ( figure [ fig_rvalue ] ) .
note that @xmath79 is not very sensitive to the temperature . with the definition of @xmath79 given above
, the final departure coefficient can be written as @xcite : @xmath87_{lte}}{1+r \left[n^+_{3/2}/n^+_{1/2}\right]_{lte}}.\end{aligned}\ ] ] the final departure coefficient is the value that we are interested in to describe crrls .
value as a function of electron temperature , in a range of densities .
the @xmath79 value is nearly independent of temperature , and for @xmath88 , @xmath89 .
right panel : ion `` lte '' ratios as a function of @xmath27 , independent of density.[fig_rvalue],title="fig : " ] value as a function of electron temperature , in a range of densities .
the @xmath79 value is nearly independent of temperature , and for @xmath88 , @xmath89 .
right panel : ion `` lte '' ratios as a function of @xmath27 , independent of density.[fig_rvalue],title="fig : " ] having described how to derive the @xmath90 , now we focus on the problem of obtaining the departure coefficients for both ion cores from the level population equation .
we use the same procedure to obtain the departure coefficients for both parent ion cores , as the only difference in the level population equation for the @xmath13 and the @xmath12 cores is the inclusion of dielectronic recombination and autoionization processes .
we will refer as @xmath40 and @xmath42 without making a distinction between the @xmath13 and @xmath12 in this subsection .
we follow the methods described in @xcite and improved in @xcite to solve the level population equation in an iterative manner .
first , we solve the level population equation by assuming that the @xmath35 sublevels are in statistical equilibrium , i.e. @xmath91 for all @xmath35 sublevels .
we refer to this approach as the @xmath14-method ( see appendix b ) .
second , we used the previously computed values to determine the coefficients on the right hand side of equation [ eqn_llevelpopc ] that contain terms with @xmath92 .
thus , the level population equation for a given @xmath14 is a tridiagonal equation on the @xmath35 sublevels involving terms of the type @xmath93 .
this tridiagonal equation is solved for the @xmath40 values ( further details are given in appendix b ) .
the second step of this procedure is repeated until the difference between the computed departure coefficients is less than 1% .
we consider a fixed maximum number of levels , @xmath94 , equal to 9900 .
we make no explicit assumptions on the asymptotic behavior of the @xmath42 for larger values of @xmath14 .
therefore , no fitting or extrapolation is required for large @xmath14 .
the adopted value for @xmath94 is large enough for the asymptotic limit
@xmath95 for @xmath96 to hold even at the lowest densities considered here . for the @xmath49-method
, we need to consider all @xmath35 sublevels up to a high level ( @xmath97 ) . for levels higher than this critical @xmath14 level ( @xmath98 ) ,
we assume that the @xmath35 sublevels are in statistical equilibrium . in our calculations @xmath99 , regardless of the density . in this section , we provide a brief description of the rates used in solving the level populations .
further details and the mathematical formulations for each rate are given in appendices c , d , e and f. accurate values for the rates are critical to obtain meaningful departure coefficients when solving the level population equation ( equation [ eqn_llevelpopc ] ) .
radiative rates are known to high accuracy ( @xmath100 ) as they can be computed from first principles . on the other hand ,
collision rates at low temperatures are more uncertain ( @xmath101 , @xcite ) .
the einstein coefficients for spontaneous and stimulated transitions can be derived from first principles .
we used the recursion formula described in @xcite to obtain the values for the einstein @xmath57 coefficients . to solve the @xmath14 method
( our first step in solving the level population equation ) we require the values for @xmath102 , which can be easily obtained by summing the @xmath57 : @xmath103 the mathematical formulation to obtain values for spontaneous transitions is detailed in appendix [ app_einsa ] .
the coefficients for stimulated emission and absorption ( @xmath104 ) are related to the @xmath102 coefficients by : @xmath105 in general , energy changing collisions are dominated by the interactions of electrons with the atom .
the interaction of an electron with an atom can induce transitions of the type : @xmath106 with @xmath92 changing the distribution of electrons in an atom population .
@xcite used the formulation of @xcite .
the collision rates derived by @xcite are essentially the same as @xcite
. however , the collision rates from @xcite are not valid for the low temperatures of interest here .
instead , we use collision rates from @xcite .
we note that at high @xmath27 and for high @xmath14 levels , the bethe ( born ) approximation holds and values of the rates from @xcite differ by less than 20% when compared to those from @xcite .
the good agreement between the two rates is expected since the results from @xcite are based on @xcite . on the other hand , at low @xmath27 and for low @xmath14 levels values
the two rates differ by several orders of magnitude and , indeed , the @xcite values are too high to be physically realistic . a comparison of the rates for different values of @xmath27 and @xmath107 transitions is shown in figure [ fig_col ] .
we explore the effects of using @xcite rates on the @xmath42 values in section [ section_comparison ] .
the inverse rates are obtained from detailed balance : @xmath108 in order to solve the @xmath49-method , rates of the type @xmath60 with @xmath61 are needed . here
, the approach of @xcite is followed and the collision rates are normalized by the oscillator strength of the transitions ( equation 5 in @xcite ) . only transitions with @xmath109 were included as these dominate the collision process @xcite , . and at low @xmath14 levels .
as is well known , transitions with @xmath110 dominate .
the difference between @xmath111 and @xmath110 rates is less at lower @xmath27.[fig_col],title="fig : " ] and at low @xmath14 levels .
as is well known , transitions with @xmath110 dominate .
the difference between @xmath111 and @xmath110 rates is less at lower @xmath27.[fig_col],title="fig : " ] and at low @xmath14 levels .
as is well known , transitions with @xmath110 dominate .
the difference between @xmath111 and @xmath110 rates is less at lower @xmath27.[fig_col],title="fig : " ] and at low @xmath14 levels .
as is well known , transitions with @xmath110 dominate .
the difference between @xmath111 and @xmath110 rates is less at lower @xmath27.[fig_col],title="fig : " ] for low @xmath14 levels , the @xmath35 level population has to be explicitly calculated .
moreover , for the dielectronic recombination process , the angular momentum changing collisions set the value for which the dielectronic recombination process is important , and transitions of the type : @xmath112 must be considered . in general ,
collisions with ions are more important than collisions with electrons . here ,
for simplicity , we adopt that @xmath113 is the dominant cation .
@xcite used @xmath35-changing collision rates from @xcite which are computed iteratively for a given @xmath14 level starting at @xmath114 or @xmath115 .
however , as pointed out by @xcite and @xcite , the values for the @xmath35-changing rates obtained by starting the iterations at @xmath114 differ from those obtained when starting at @xmath115 .
moreover , averaging the @xmath35-changing rates obtained by the two different initial conditions leads to an oscillatory behavior of the rates that depends on @xmath35 @xcite .
@xcite circumvented this problem by normalizing the value of the rates by the oscillator strength ( equation 4 in @xcite ) .
in addition , at high @xmath14 levels and high densities the values for @xmath60 can become negative ( equation 43 in @xcite ) .
this poses a problem when studying the level population of carbon atoms at the high @xmath14 levels of interest in the present work levels for high @xmath14 . ] .
the more recent study of @xcite provides a general formulation to obtain the value of @xmath35-changing transition rates .
these new rates use a much smaller cut - off radius of the probability of the transition for large impact parameters .
furthermore , the rates from @xcite are well behaved over a large range of temperature and densities and they do not exhibit the oscillatory behavior with @xmath35 sublevel shown by the @xcite rates .
therefore , we use the @xcite rates in this work . @xcite derived the following expression , valid for @xmath116 and @xmath117 : @xmath118.\end{aligned}\ ] ] where @xmath119 is the bohr radius and @xmath120 is the reduced mass of the system .
values for the inverse process are obtained by using detailed balance : @xmath121 we note that the @xmath35-changing collision rates obtained by using the formula from @xcite can differ by a factor of six @xcite with those using the @xcite formulation .
we discuss the effect on the final @xmath42 values in section [ section_comparison ] , where we compare our results with those of @xcite in the hydrogenic approximation and with those of @xcite for carbon atoms .
radiative ionization occurs when an excited atom absorbs a photon with enough energy to ionize the excited electron .
the process can be represented as follows : @xmath122 and the inverse process is radiative recombination .
we use the recursion relation described in @xcite to obtain values for the ionization cross - section ( appendix [ app_radrec ] ) .
values for the radiative recombination ( @xmath65 ) coefficients were obtained using the milne relation and standard formulas ( e.g. @xcite , appendix [ app_radrec ] ) .
the program provided by @xcite only produces reliable values up to @xmath123 due to cancellation effects in the iterative procedure . in order to avoid cancellation effects ,
the values computed here were obtained by working with logarithmic values in the recursion formula . as expected ,
our values for the rates match those of @xcite well . for the @xmath14-method
we require the sum of the individual @xmath65 values : @xmath124 the averaged @xmath125 values agree well with the approximated formulation of @xcite to better than 5% , validating our approach .
collisional ionization occurs when an atom encounters an electron and , due to the interaction , a bound electron from the atom is ionized .
schematically the process can be represented as : @xmath126 the inverse process is given by the 3-body recombination and the value for the 3-body recombination rate is obtained from detailed balance : @xmath127 we used the formulation of @xcite and compared the values with those from the formulation given by @xcite . for levels above 100 and at @xmath128 , the brocklehurst & salem values are a factor of @xmath129 larger , but the differences quickly decrease for higher temperatures . to obtain the @xmath63 that are needed in the @xmath49-method
, we followed @xcite and assumed that the rates are independent of the angular momentum .
the mathematical formulation is reproduced in the appendix [ app_cni ] for convenience of the reader .
the dielectronic recombination process involves an electron recombining into a level @xmath14 while simultaneously exciting one of the bound electrons ( left side of equation [ dielrec1 ] , below ) .
this state ( @xmath130 ) is known as an autoionizing state . in this autoionizing state
, the atom can stabilize either by releasing the recombined electron through autoionization ( inverse process of dielectronic recombination ) or through radiative stabilization ( right hand side of equation [ dielrec1 ] ) .
dielectronic recombination and autoionization are only relevant for atoms with more than one electron .
@xmath131 for c@xmath53 recombination , at @xmath132 free electrons in the plasma can recombine to a high @xmath14 level , and the kinetic energy is transfered to the core of the ion , producing an excitation of the @xmath11 fine - structure level of the c@xmath53 atom core ( which has a difference in energy @xmath85 ) . due to the long radiative lifetime of the fine - structure transition ( @xmath133 ) , radiative stabilization can be neglected . following @xcite , who compute the autoionization rate using the formulation by @xcite , i. e. : @xmath134 with @xmath135 the collision strength for the @xmath11 excitation at the threshold .
as @xcite , we used the formula obtained by @xcite : @xmath136 valid for @xmath137 . in order to avoid the singularity at @xmath114 we computed the autoionization rate , @xmath55 , from the approximate expression given in @xcite : @xmath138 which is valid for @xmath139
the dielectronic recombination rate is obtained by detailed balance : @xmath140 @xcite defined @xmath141 as the departure coefficient when autoionization / dielectronic recombination dominate : @xmath142,\nonumber\\ & = & \frac{1}{r}.\end{aligned}\ ] ] a comparison of the dielectronic recombination rate with values from the literature is hampered by the focus of previous works on higher temperatures .
the calculations of dielectronic recombination rates for carbon from @xcite did not include fine structure transitions and are not suited for a direct comparison with the study presented here .
furthermore , the values presented by @xcite are given for higher temperatures than those studied here .
the more recent study of @xcite provides state resolved values for dielectronic recombination rates . for the physical conditions of interest in this article and for the fine structure levels of interest here ,
@xcite provides values using intermediate coupling .
however , a direct comparison with the results from @xcite is not possible since they only include @xmath143 type dielectronic recombination resonances associated with the excitation of a @xmath144 electron to a @xmath145 level ( see equation 1 in @xcite ) .
the behavior of crrls with frequency depends on the level population of carbon via the departure coefficients .
we compute departure coefficients for carbon atoms by solving the level population equation using the rates described in section [ section_rates ] and the approach in section [ section_nummethod ] . here , we present values for the departure coefficients and provide a comparison with earlier studies in order to illustrate the effect of our improved rates and numerical approach .
a detailed analysis of the line strength under different physical conditions relevant for the diffuse clouds and the effects of radiative transfer are provided in an accompanying article ( paper ii ) .
the final departure coefficients for carbon atoms ( i.e. @xmath146 ) are obtained by computing the departure coefficients recombining from both parent ions , those in the @xmath12 level and those in the @xmath13 level .
therefore , it is illustrative to study the individual departure coefficients for the @xmath12 core , which are hydrogenic , and the departure coefficients for the @xmath13 core separately . in figure [ fig_bn1d4 ]
we show example @xmath42 and @xmath147 values obtained in the hydrogenic approximation at @xmath148 for a large range in density . the behavior of the @xmath42 values as a function of @xmath14 can be understood in terms of the rates that are included in the level population equation . at the highest @xmath14 levels , collisional ionization and three body recombination
dominate the rates in the level population equation and the @xmath42 values are close to unity .
we can see that as the density increases , collisional equilibrium occurs at lower @xmath14 levels and the @xmath42 values approach unity at lower levels . in contrast , for the lowest @xmath14 levels , the level population equation is dominated by radiative processes and the levels drop out of collisional equilibrium . as the radiative rates increase with decreasing @xmath14 level ,
the departure coefficients become smaller .
we note that differences in the departure coefficients for the low @xmath14 levels for different temperatures are due to the radiative recombination rate , which has a @xmath149 dependence . at intermediate @xmath14 levels , the behavior of the @xmath42 as a function of @xmath14 shows a more complex pattern with a pronounced `` bump '' in the @xmath42 values for intermediate levels ( @xmath150 ) .
to guide the discussion we refer the reader to figure [ fig_bn1d4 ] .
starting at the highest @xmath14 , @xmath151 , as mentioned above .
for these high @xmath14 levels , @xmath35-changing collisions efficiently redistribute the electron population among the @xmath35 states and , at high density , the @xmath40 departure coefficients are unity as well ( figure [ fig_bnl ] , upper panels ) .
for lower values of @xmath14 , the @xmath42 values decrease due to an increased importance of spontaneous transitions . at these @xmath14 levels , the @xmath42 values obtained by the @xmath49 method differ little from the values obtained by the @xmath14-method , since @xmath35-changing collisions efficiently redistribute the electrons among the @xmath35 sublevels for a given @xmath14 level . for lower @xmath14 levels , the effects of considering the @xmath35 sublevel distribution become important as @xmath35-changing collisions compete with spontaneous decay , effectively `` storing '' electrons in high @xmath35 sublevels for which radiative decay is less important .
collisions compete with spontaneous decay , effectively `` storing '' electrons in high @xmath35 sublevels for which radiative decay is less important . specifically , the spontaneous rate out of a given level is approximately @xmath152 , and is higher for lower @xmath35 sublevels .
thus , high-@xmath35 sublevels are depopulated more slowly relative to lower @xmath35 sublevels on the same @xmath14 level .
this results in a slight increase in the departure coefficients . reflecting the statistical weight factor in equation [ eqn_bn2 ]
, the higher @xmath35 sublevels dominate the final @xmath42 value resulting in an increase in the final @xmath42 value . as the density increases , the @xmath35 sublevels approach statistical distribution faster . as a result ,
the influence of the @xmath35 sublevel population on the final @xmath42 is larger for lower densities than for higher densities at a given @xmath27 .
the interplay of the rates produce the `` bump '' which is apparent in the @xmath42 distribution ( figure [ fig_bn1d4 ] ) the influence of @xmath35-changing collisions on the level populations and the resulting increase in the @xmath42 values was already presented by @xcite and analyzed in detail by @xcite in the context of hydrogen masers .
the results of our level population models are in good agreement with those provided by @xcite as we show in section [ section_comparison ] .
values ( left ) and @xmath147 values ( right ) for hydrogen at @xmath153 ( upper and lower panels , respectively ) for different densities ( @xmath28 , colorscale ) .
the departure coefficients obtained using the @xmath49-method show a `` bump '' at low @xmath14 levels .
the strength and position of the `` bump '' depend on the physical conditions . as density increases ,
the @xmath35-changing collisions redistribute the electron population more effectively.[fig_bn1d4 ] ] values at low densities ( _ left panel _ ) and high densities ( _ right panel _ ) .
statistical distribution of the @xmath35-sublevels is attained at levels as low as @xmath154 . for lower levels ,
radiative processes dominate the level population . at low density ( _ left panel _ ) , radiative processes dominate even at high @xmath14 levels.[fig_bnl ] ] only carbon atoms recombining to the @xmath13 ion core are affected by dielectronic recombination .
having analyzed the departure coefficients for the hydrogenic case , we focus now on the @xmath67 values and the resulting @xmath146 as introduced in section [ section_levelpop ] .
figure [ fig_bn32carbon ] show example values for @xmath69 for @xmath155 and electron densities between @xmath156 .
as pointed out by @xcite , the low lying @xmath35 sublevels are dominated by the dielectronic process and the @xmath67 values are equal to @xmath141 ( equation [ eqn_bdi ] ) . as can be seen in figure [ fig_rvalue ] , such values can be much larger than unity at low densities resulting in an overpopulation of the low @xmath14 levels for the @xmath157 ion cores . in figure [ fig_bnfinalcarbon ]
we show @xmath146 as a function of @xmath14 level under the same conditions .
we see that at high electron densities the departure coefficients show a similar behavior as the hydrogenic values .
furthermore , an increase in the level population to values larger than unity is seen at low densities and moderate to high temperatures . to guide the discussion
, we analyze the behavior of the @xmath146 when autoionization / dielectronic recombination dominates .
this occurs at different levels depending on the values of @xmath27 and @xmath28 considered .
nevertheless , it is instructive to understand the behavior of the level population in extreme cases .
when autoionization / dielectronic recombination dominate , the @xmath146 in equation [ equation_bnfinal2 ] is given by : @xmath158_{lte}}{1+r \left[n^+_{3/2}/n^+_{1/2}\right]_{lte}}.\end{aligned}\ ] ] at high densities , @xmath79 approaches unity and we note two cases .
the first case is when @xmath27 is high , the maximum value of @xmath159_{lte}=2 $ ] , meaning that a large fraction of the ions are in the @xmath13 core .
consequently , @xmath160 , thus the effect of dielectronic recombination is to increase the level population as compared to the hydrogenic case .
we also note that since @xmath161 the final @xmath162 . the second case we analyze is for low @xmath27 , where the ion lte ratio is low and most of the ions are in the @xmath12 core .
thus , @xmath163 and the departure coefficients are close to hydrogenic . at low densities , @xmath164 and , as above
, we study two cases .
the first is when @xmath27 is high , the maximum value of @xmath159_{lte}=2 $ ] and @xmath165 , therefore dielectronic recombination produces a large overpopulation as compared to the hydrogenic case .
the second case is when @xmath27 is low and most of the ions are in the @xmath12 level and , as in the high density case , the @xmath163 .
we note from this analysis that overpopulation of the @xmath146 ( relative to the hydrogenic case ) is only possible for a range of temperatures and densities .
in particular , @xmath146 is maximum for high temperatures and low densities .
having analyzed the behavior of the @xmath146 values in the extreme @xmath166 case , now we analyze the behavior of @xmath69 with @xmath14 .
the population in the low @xmath14 levels is dominated by dielectronic recombination @xcite and @xmath166 up until a certain @xmath14 level where @xmath69 begins to decrease down to a value of one .
the @xmath14 value where this change happens depends on temperature , moving to higher @xmath14 levels as @xmath27 decreases . to understand this further
, we analyze the rates involved in the @xmath35 sublevel population ( figure [ fig_bn32carbon ] ) .
the low @xmath35 sublevels are dominated by dielectronic recombination and autoionization and the @xmath40 values for the @xmath157 ion cores are @xmath167 . for the higher @xmath35 sublevels other processes ( mainly collisions )
populate or depopulate electrons from the level @xmath14 and the net rate is lower than that of the low @xmath35 dielectronic recombination / autoionization .
this lowers the @xmath40 value , which is effectively delayed by @xmath35-changing collisions since they redistribute the population of electrons in the @xmath14 level .
the @xmath40 for highest @xmath35 values dominate the value of @xmath69 due to the statistical weight factor .
we note that the behavior of the @xmath69 cores as a function of @xmath14 ( see figure [ fig_exbn32 ] ) can be approximated by : @xmath168 ^ 3 \right)\times \left(b_{di}-1\right ) + 1.\end{aligned}\ ] ] with @xmath141 defined as in @xcite ( equation [ eqn_bdi ] ) and @xmath169 was derived from fitting our results : @xmath170 in diffuse clouds the integrated line to continuum ratio is proportional to @xmath147 .
we note that the @xmath45 behavior is more complex as can be seen in figure [ fig_bnbetacarbon ] .
the low @xmath14 `` bump '' on the @xmath146 makes the @xmath147 high at low densities and for levels between about @xmath171 and @xmath172 .
since the @xmath146 values decrease from values larger than one to approximately one , the @xmath45 changes sign . in figure [ fig_bnbetanzero ]
we show the electron density as a function of the level where the change of sign on the @xmath147 occurs . at temperatures higher than about @xmath173 , our models for @xmath174 show no change of sign due to the combined effects of @xmath35-changing collisions and dielectronic recombination .
parent ions as a function of @xmath14 at @xmath175 for different densities ( @xmath28 , colorscale ) .
the values for low @xmath14 levels are close to @xmath141 and decrease towards a value of one . at high densities ,
@xmath176.[fig_bn32carbon ] ] ) as a function of @xmath14 level at @xmath175 for different densities ( @xmath28 , colorscale ) .
the `` bump '' seen in hydrogenic atoms is amplified by dielectronic recombination .
as density increases the @xmath146 are closer to the hydrogenic value.[fig_bnfinalcarbon ] ] values for carbon as a black line ( solid ) , the discontinuity at @xmath177 is due to the @xmath98 value .
overplotted as a red ( dot dashed ) line is the approximation in equation [ eqn_appbn32 ] .
the blue ( dashed ) line is the value of @xmath141.[fig_exbn32 ] ] values for carbon atoms at @xmath175 for different densities
( @xmath28 , colorscale).[fig_bnbetacarbon ] ] values go to zero for @xmath178 . at temperatures larger than 200 k and for an electron densities around @xmath179
the @xmath180 values do not go through zero.[fig_bnbetanzero ] ] the level population of hydrogenic atoms is a well studied problem . here , we will describe the effects of the updated collision rates as well as point out differences due to the improved numerical method . at the lowest densities , we can compare our results for hydrogenic atoms with the values of @xcite for hydrogen atoms .
the results of @xcite were obtained in the low density limit , i.e. no collision processes were taken into account in his computations .
the results are given in terms of the emissivity of the line normalized by the h@xmath181 emissivity .
as can be seen in figure [ fig_compmartin ] , our results agree to better than 5% , and for most levels to better than 0.5% . at high densities
, we compare the hydrogenic results obtained here with those of @xcite .
our approach reproduces well the @xmath40 ( and @xmath42 ) values of @xcite ( to better than 1% ) when using the same collision rates @xcite as can be seen in figure [ fig_effofcol ] .
we note that the effect of using different energy changing rates ( @xmath182 ) has virtually no effect on the final @xmath183 values . on the other hand , using @xcite values for the @xmath184 rates results in differences in the @xmath42 values of 30% at @xmath185 . as expected , the difference is less at higher temperatures and densities since values are closer to equilibrium ( see figure [ fig_comphs ] ) . at low @xmath14 levels ,
our results for high @xmath35 levels are overpopulated as compared to the values of @xcite leading to an increases in the @xmath42 values . ) for low @xmath49 lines at low density and the results from @xcite in the low @xmath28 approximation .
our results agree to better than 1% at most levels.[fig_compmartin],title="fig : " ] ) for low @xmath49 lines at low density and the results from @xcite in the low @xmath28 approximation .
our results agree to better than 1% at most levels.[fig_compmartin],title="fig : " ] values for @xmath186 and @xmath187 .
h&s are the departure coefficients from @xcite who used @xcite ; gplr corresponds to the use of @xmath14 changing collision rates from @xcite , v&s from @xcite ; p&s corresponds to the use of @xmath35-changing collision rates from @xcite , vos corresponds to @xcite .
the largest differences are @xmath188 due to the use of different @xmath35-changing collision rates.[fig_effofcol],title="fig : " ] values for @xmath186 and @xmath187 .
h&s are the departure coefficients from @xcite who used @xcite ; gplr corresponds to the use of @xmath14 changing collision rates from @xcite , v&s from @xcite ; p&s corresponds to the use of @xmath35-changing collision rates from @xcite , vos corresponds to @xcite .
the largest differences are @xmath188 due to the use of different @xmath35-changing collision rates.[fig_effofcol],title="fig : " ] values ( black ) at @xmath189 and @xmath190 and the results from @xcite ( red line , dashed ) .
differences are due to the use of @xmath35-changing collision rates from @xcite.[fig_comphs ] ] now we compare departure coefficients obtained here with the results of @xcite and the effect of including @xmath35-changing collisions on the departure coefficients , see figures [ fig_compps ] and [ fig_compps2 ] .
we will focus the discussion on the @xmath42 values from @xcite as the @xcite values are similar . while the results presented here are remarkably different from those of @xcite and @xcite , some trends are similar .
we will first discuss the differences .
our results in figures [ fig_compps ] and [ fig_compps2 ] show a pronounced bump for low @xmath14 in the range 50 to 150 .
this bump is similar to what we see for the hydrogenic approximation but enhanced by dielectronic recombination ( c.f .
figures 3.6 and 8 ; section 3.1.2 ) .
as discussed in section 3.1.1 this bump arises at these intermediate @xmath14 levels because collisions compete with spontaneous decay , effectively storing electrons in high @xmath35 sublevels for which radiative decay is less important .
this means that the inclusion of @xmath35-changing collisions leads to significantly larger @xmath183 values for @xmath14 in the range 50 to 150 as compared to @xcite .
regardless of the @xmath35-changing collision rates used , at higher @xmath14 we note that our @xmath42 values with increasing @xmath14 asymptotically approach unity much faster than @xcite .
this is especially true for lower electron densities ( @xmath1911.0 @xmath192 ) and a direct consequence of using the @xmath49-method to compute the departure coefficients .
although the detailed behavior of our @xmath183 values differs strongly from @xcite there are also similarities in the general trends that we observe as a function of electron density and temperature . in particular , the very low and very high @xmath14 asymptotic behavior of the @xmath183 values is similar to @xcite in that the highest electron densities for a given electron temperature have the lowest @xmath183 values at low @xmath14 and approach equilibrium ( @xmath42=1 ) the fastest with increasing @xmath14 . for higher electron densities and lower electron temperatures , our results become increasingly similar to the hydrogenic case and agree with @xcite .
this is expected as , as discussed in section 3.1 at high densities the @xmath42 values approach equilibrium . in terms of @xmath180
our results show , as expected , good agreement with the hydrogenic case and @xcite in the high density and low temperature limit .
however , for the lower densities and higher temperatures shown in figures [ fig_compps ] and [ fig_compps2 ] our models predict @xmath180 values that are lower by up to about an order of magnitude as compared to @xcite .
this is particularly striking for the @xmath27=100 k and @xmath193=0.05 @xmath192 model shown in figure [ fig_compps2 ] where we find that both the maximum negative @xmath180 value and maximum positive @xmath180 value are more than an order of magnitude lower than the corresponding @xcite values .
since the integrated optical depth is directly proportional to the value of @xmath180 ( e.g. @xcite ) we can interpret @xmath180 as a stimulation factor .
this means that , for a given set of physical conditions , our models predict much lower maximum integrated optical depths for carbon as compared to earlier investigations ( e.g. @xcite ) .
this is true for both emission ( negative @xmath180 ) and absorption ( positive @xmath180 ) .
in particular , our models predict that equilibrium will be reached at much lower @xmath14 ( typically around @xmath194 ) and thus that the integrated optical at high @xmath14 ( low frequencies ) will show a rather flat behavior for @xmath195600 whereas the previous models by @xcite and @xcite predict a strong increase with increasing @xmath14 .
we find that although our @xmath42 values asymptotically approach equilibrium at high @xmath14 that this value is not yet reached at @xmath196 .
therefore , the @xmath180 values we find are nearly , but not yet completely , constant in the range @xmath197600 - 1000 and as such the dependence of integrated optical depth on @xmath180 remains important at high @xmath14 .
finally , we note that for sufficiently high electron temperatures and low electron densities our models predict the existence of a region at intermediate @xmath14 ( @xmath197100 - 200 ) where the @xmath180 values can become positive . this behavior is a direct consequence of the inclusion of @xmath35-changing collisions in our models .
a more detailed comparison of the departure coefficients obtained using the @xmath35-changing collision from @xcite and those using the rates from @xcite ( figure [ fig_comppengelly ] ) reveals differences of less than 30% for the conditions of interest for crrl studies .
apart from the @xmath35-changing collisions there are other potentially important differences between our models and those published by @xcite .
@xcite do not provide the explicit values of the dielectronic recombination rates that they use .
however , they refer back to @xcite for these rates and as we use the same formalism , we do not think that the dielectronic recombination rates are at the heart of the discrepancy . may have influenced their results .
in addition , we note that we use somewhat different collision rates in our simulations .
however , as illustrated in figure 13 and 14 , the exact collision rates have only limited influence on the @xmath42 values .
rather , we suspect that the approximate way the statistical equilibrium equations are solved by @xcite may have influenced their results and that including @xmath35-changing collisions properly rather than adopting a statistical populations as did @xcite is key
. a further assessment of the effect of any uncertainty in the adopted dielectronic recombination rates on the final departure coefficients can be performed by arbitrarily multiplying the dielectronic recombination rate by a factor .
we note that a dielectronic recombination rate a factor of 30% higher ( lower ) increases ( decreases ) the departure coefficients at low levels ( @xmath198 ) by 30% . at the higher levels of interest for the study of crrls , ( @xmath199 ) a factor of 30% on the dielectronic recombination rates
changes the values of the departure coefficients by less than 10% ( figure [ fig_effectofdielec ] , upper panels ) .
as expected , the values for @xmath180 are affected more by the change on the dielectronic recombination rate and can be altered by a factors of a few ( figure [ fig_effectofdielec ] , lower panels ) .
it is clear that quantitative interpretation of carbon radio recombination lines would be served by more accurate dielectronic recombination rates that include the fine structure levels .
-changing collision rates from @xcite ( panels b ) and e ) ) and those from @xcite ( panels c ) and f ) ) at @xmath27=50 k. lines marked as 1 , 2 , 4 correspond to electron densities @xmath2000.05 , 0.1 and 1.0 @xmath192 ( solid , dotted and dashed lines ) respectively .
the top panels show @xmath42 vs. @xmath14 and the bottom panels show the product @xmath180 vs. @xmath14.[fig_compps],title="fig : " ] -changing collision rates from @xcite ( panels b ) and e ) ) and those from @xcite ( panels c ) and f ) ) at @xmath27=50 k. lines marked as 1 , 2 , 4 correspond to electron densities @xmath2000.05 , 0.1 and 1.0 @xmath192 ( solid , dotted and dashed lines ) respectively .
the top panels show @xmath42 vs. @xmath14 and the bottom panels show the product @xmath180 vs. @xmath14.[fig_compps],title="fig : " ] -changing collision rates from @xcite ( panels b ) and e ) ) and those from @xcite ( panels c ) and f ) ) at @xmath27=50 k. lines marked as 1 , 2 , 4 correspond to electron densities @xmath2000.05 , 0.1 and 1.0 @xmath192 ( solid , dotted and dashed lines ) respectively .
the top panels show @xmath42 vs. @xmath14 and the bottom panels show the product @xmath180 vs. @xmath14.[fig_compps],title="fig : " ] ( 10,10 ) ( 36,270)a ) ( 186,270)b ) ( 346,270)c ) ( 36,125)d ) ( 186,125)e ) ( 346,125)f ) -changing collision rates from @xcite ( panels b ) and e ) ) and those from @xcite ( panels c ) and f ) ) at @xmath27=100 k. lines marked as 1 , 2 , 4 correspond to electron densities @xmath2000.05 , 0.1 and 1.0 @xmath192 ( solid , dotted and dashed lines ) respectively .
the top panels show @xmath42 vs. @xmath14 and the bottom panels show the product @xmath180 vs. @xmath14.[fig_compps2],title="fig : " ] -changing collision rates from @xcite ( panels b ) and e ) ) and those from @xcite ( panels c ) and f ) ) at @xmath27=100 k. lines marked as 1 , 2 , 4 correspond to electron densities @xmath2000.05 , 0.1 and 1.0 @xmath192 ( solid , dotted and dashed lines ) respectively .
the top panels show @xmath42 vs. @xmath14 and the bottom panels show the product @xmath180 vs. @xmath14.[fig_compps2],title="fig : " ] -changing collision rates from @xcite ( panels b ) and e ) ) and those from @xcite ( panels c ) and f ) ) at @xmath27=100 k. lines marked as 1 , 2 , 4 correspond to electron densities @xmath2000.05 , 0.1 and 1.0 @xmath192 ( solid , dotted and dashed lines ) respectively .
the top panels show @xmath42 vs. @xmath14 and the bottom panels show the product @xmath180 vs. @xmath14.[fig_compps2],title="fig : " ] ( 10,10 ) ( 36,270)a ) ( 186,270)b ) ( 346,270)c ) ( 36,125)d ) ( 186,125)e ) ( 346,125)f ) ( 100,200 ) ( 0,0)-changing collisions rates from @xcite and @xcite at 50 k and 100 k. the largest differences are @xmath201 at levels @xmath202.[fig_comppengelly],title="fig : " ] -changing collisions rates from @xcite and @xcite at 50 k and 100 k. the largest differences are @xmath201 at levels @xmath202.[fig_comppengelly],title="fig : " ] ( 55,135)a ) ( 295,135)b ) ( left ) and 0.1 @xmath203 ( right ) .
lower panels : @xmath147 for the same physical conditions as in the upper panels . at @xmath14 levels
lower than @xmath204 the @xmath147 values derived by using the modified dielectronic recombination rates ( blue and red ) are similar to those without modification ( black ) . at higher levels the overall trends are similar but they can differ by factors of a few.[fig_effectofdielec ] ]
we have solved the level population equation for hydrogenic atoms using novel rates involved in the process .
the level population equation is solved in two approximations : the @xmath14 and the @xmath49 method . the departure coefficients obtained using the @xmath14 method are similar to values from the literature ( e.g. @xcite and @xcite ) .
our results using the @xmath49 method reproduce those from @xcite well , once allowance is made for updates in the collisional rates . by including the dielectronic recombination process together with the @xmath49 method we are able to model the level population of carbon in terms of the departure coefficients .
our results are qualitatively similar to those of @xcite .
however , the values obtained here differ considerably from those from the literature .
the differences can be understood in terms of the use of improved collision rates and the improved numerical approach using the @xmath49 method .
we confirm that dielectronic recombination can indeed produce an increase on the values of the departure coefficients at high @xmath14 levels compared to the hydrogenic values . in anticipation of low frequency radio recombination line surveys of the diffuse interstellar
medium now being undertaken by lofar , we have expanded the range of applicability of the formulation to the conditions of the cold neutral medium .
for this environment , external radiation fields also become important at intermediate principal quantum levels while at high levels the influence of radiation fields on the level population is less important in an accompanying paper @xcite , we discuss the expected line strength for low frequency carbon radio recombination lines and the influence of an external radiation field . throughout this work
we have used a zero radiation field . in this companion paper
we compare our results to existing observations of crrls towards cas a and regions in the inner galaxy .
we also describe the analysis techniques and diagnostic diagrams that can be used to analyze the forthcoming lofar crrl survey .
the departure coefficients obtained here will be used to analyze the lofar observations of cas a in a future article @xcite .
allen , c. w. 1973 , london : university of london , athlone press , |c1973 , 3rd ed .
altun , z. , yumak , a. , badnell , n. r. , colgan , j. , & pindzola , m. s. 2004 , , 420 , 775 asgekar , a. , oonk , j. b. r. , yatawatta , s. , et al .
2013 , , 551 , ll11 baker , j. g. , & menzel , d. h. 1938 , , 88 , 52 barinovs , g. , van hemert , m.c . , krems , r. , dalgarno , a. 2005 , , 620 , 537 brocklehurst , m. 1970 , , 148 , 417 brocklehurst , m. 1971 , , 153 , 471 brocklehurst , m. , & seaton , m. j. 1972 , , 157 , 179 brocklehurst , m. 1973 , , 14 , 81 brocklehurst , m. , & salem , m. 1975 , computer physics communications , 9 , 258 brocklehurst , m. , & salem , m. 1977 , computer physics communications , 13 , 39 burgess , a. 1958 , , 118 , 477 burgess , a. 1965 , , 69 , 1 burgess , a. , & percival , i. c. 1968 , advances in atomic and molecular physics , 4 , 109 cox , d. p. 2005
, , 43 , 337 dickinson , a. s. 1981 , , 100 , 302 dupree , a. k. 1972 , , 173 , 293 ellingson , s. w. , taylor , g. b. , craig , j. , et al .
2013 , ieee transactions on antennas and propagation , 61 , 2540 elmegreen , b. g. , & scalo , j. 2004 , , 42 , 211 erickson , w. c. , mcconnell , d. , & anantharamaiah , k. r. 1995 , , 454 , 125 ferrire , k. m. 2001 , reviews of modern physics , 73 , 1031 field , g. b. , goldsmith , d. w. , & habing , h. j. 1969 , , 155 , l149 flower , d.r .
1988 , j phys b , 21 , l451 gee , c. s. , percival , l. c. , lodge , j. g. , & richards , d. 1976 , , 175 , 209 goldberg , l. 1966 , , 144 , 1225 gordon , m. a. , & sorochenko , r. l. 2009 , astrophysics and space science library , 282 , gu , m. f. 2003 , , 590 , 1131 pengelly , r. m. , & seaton , m. j. 1964 , , 127 , 165 hayes , m. a. , & nussbaumer , h. 1984 , , 134 , 193 heiles , c. , & troland , t. h. 2003 , , 145 , 329 heiles , c. , & troland , t. h. 2003 , , 586 , 1067 hilborn , r. c. 1982 , american journal of physics , 50 , 982 hummer , d. g. , & storey , p. j. 1987 , , 224 , 801 hummer , d. g. , & storey , p. j. 1992 , , 254 , 277 jacobs , v. l. , & davis , j. 1978 , , 18 , 697 kalberla , p. m. w. , burton , w. b. , hartmann , d. , et al .
2005 , , 440 , 775 kalberla , p. m. w. , & kerp , j. 2009 , , 47 , 27 kantharia , n. g. , anantharamaiah , k. r. , & payne , h. e. 1998 , , 506 , 758 konovalenko , a. a. , & sodin , l. g. 1980 , , 283 , 360 konovalenko , a. a. , & sodin , l. g. 1981 , , 294 , 135 kulkarni , s. r. , & heiles , c. 1987 , interstellar processes , 134 , 87 launay , j .- m . , & roueff , e. 1977 , journal of physics b atomic molecular physics , 10 , 879 martin , p. g. 1988 , , 66 , 125 mckee , c. f. , & ostriker , j. p. 1977
, , 218 , 148 mckee , c. f. , & ostriker , e. c. 2007 , , 45 , 565 morabito , l. k. , van harten , g. , salgado , f. , et al .
2014 , , 441 , 2855 morabito , l. k. , oonk , j. b. r. , salgado , f. , et al .
2014 , , 795 , ll33 natta , a. , walmsley , c. m. , & tielens , a. g. g. m. 1994 , , 428 , 209 nussbaumer , h. , & storey , p. j. 1983 , , 126 , 75 oonk , j. b. r. , van weeren , r. j. , salgado , f. , et al .
2014 , , 437 , 3506 oonk , j. b. r. , morabito , l. k. , salgado , f. , et al . 2015a , arxiv:1501.01179 oonk , j. b. r. , et al , in preparation osterbrock , d. e. 1965 , , 142 , 1423 payne , h. e. , anantharamaiah , k. r. , & erickson , w. c. 1994 , , 430 , 690 percival , i. c. , & richards , d. 1978 , , 183 , 329 peters , w. m. , lazio , t. j. w. , clarke , t. e. , erickson , w. c. , & kassim , n. e. 2011 , , 525 , a128 ponomarev , v. o. , & sorochenko , r. l. 1992 , soviet astronomy letters , 18 , 215 quireza , c. , rood , r. t. , balser , d. s. , & bania , t. m. 2006 , , 165 , 338 roshi , d. a. , kantharia , n. g. , & anantharamaiah , k. r. 2002 , , 391 , 1097 rybicki , g. b. , & lightman , a. p. 1986
, radiative processes in astrophysics , by george b. rybicki , alan p. lightman , pp . 400 .
isbn 0 - 471 - 82759 - 2 .
wiley - vch , june 1986 . , salem , m. 1975 , , 173 , 513 salem , m. , & brocklehurst , m. 1979 , , 39 , 633 salgado , f. , et al . accepted ( paper ii ) savage
, b. d. , & sembach , k. r. 1996 , , 34 , 279 scalo , j. , & elmegreen , b. g. 2004 , , 42 , 275 seaton , j. m 1959 , , 119 , 81 seaton , j. m 1959 , , 119 , 90 seaton , m. j. , & storey , p. j. 1976 , atomic processes and applications .
p. g. burke ( ed . ) , north - holland publ .
co. , amsterdam , netherlands , p. 133 - 197 , 133 shaver , p. a. 1975 , pramana , 5 , 1 snow , t. p. , & mccall , b. j. 2006 , , 44 , 367 stepkin , s. v. , konovalenko , a. a. , kantharia , n. g. , & udaya shankar , n. 2007 , , 374 , 852 strelnitski , v. s. , ponomarev , v. o. , & smith , h. a. 1996 , , 470 , 1118 storey , p. j. , & hummer , d. g. 1991 , computer physics communications , 66 , 129 storey , p. j. , & hummer , d. g. 1995 , , 272 , 41 tielens , a. g. g. m. , & hollenbach , d. 1985 , , 291 , 722 tingay s. j. , goeke r. , bowman , j. d. , et al .
2013 , pasa , 30 , 7 van haarlem m. p. , wise m. w. , gunst a. w. , et al .
2013 , a&a , 556 , 2 vriens , l. , & smeets , a. h. m. 1980 , , 22 , 940 vrinceanu , d. , onofrio , r. , & sadeghpour , h. r. 2012 , , 747 , 56 walmsley , c. m. , & watson , w. d. 1982 , , 260 , 317 watson , w. d. , western , l. r. , & christensen , r. b. 1980 , , 240 , 956 weaver , h. , & williams , d. r. w. 1973 , , 8 , 1 wilson , n.j . ,
bell , k.l 2002 , , 337 , 1027 wolfire , m. g. , mckee , c. f. , hollenbach , d. , & tielens , a. g. g. m. 2003 , , 587 , 278 wyrowski , f. , schilke , p. , hofner , p. , & walmsley , c. m. 1997 , , 487 , l171 wyrowski , f. , walmsley , c. m. , goss , w. m. , & tielens , a. g. g. m. 2000 , , 543 , 245
lp100 mm + + symbol & descritpion + + symbol & + + @xmath205 & spontaneous transition rate of the carbon fine structure line @xmath13-@xmath12 + @xmath206 & autoionization rate + @xmath24 & einstein coefficient for spontaneous transition between @xmath20 and @xmath14 + @xmath207 & einstein coefficient for spontaneous transition between @xmath208 state to @xmath49 state + @xmath119 & bohr radius + @xmath209 & photoionization cross section + @xmath104 & einstein coefficient for stimulated transition from level @xmath20 to @xmath14 + @xmath42 & departure coefficient for level n + @xmath68 & departure coefficient for atoms recombining from the @xmath210 ion core for level @xmath14 + @xmath69 & departure coefficient for atoms recombining from the @xmath157 ion core for level @xmath14 + @xmath146 & departure coefficient for atoms recombining from both ion cores + @xmath211 & carbon recombination line for @xmath46 transition + @xmath212 & rates for energy changing collisions between level @xmath20 and @xmath14 + @xmath213 & coefficient for recursion relations used to obtain the radial matrices values + @xmath32 & speed of light + @xmath214 & emission measure of carbon ions + @xmath215 & statistical weight for the fine structure level @xmath13 + @xmath216 & statistical weight for the fine structure level @xmath12 + @xmath18 & planck constant + @xmath217 & intensity of the background continuum + @xmath218 & intensity of the line + @xmath219 & intensity of the continuum + @xmath220 & intensity of the fine structure line of carbon at 158 @xmath221 + @xmath15 & line emission coefficient + @xmath16 & line absorption coefficient + @xmath31 & boltzmann constant + @xmath222 & pathlength of cloud + @xmath35 & angular momentum quantum number + @xmath80 & critical density for collisions on a two level atom + @xmath223 & density of atoms in level @xmath14 + @xmath224 & density of atoms in level @xmath14 and sublevel @xmath35 + @xmath28 & electron density + @xmath225 & hydrogen density + @xmath29 & density of the parent ions + @xmath226 & level population of carbon ions in the @xmath13 core + @xmath227 & level population of carbon ions in the @xmath12 core + @xmath14 & lower principal quantum number + @xmath20 & upper principal quantum number + @xmath94 & maximum level considered in our simulations + @xmath98 & critical level considered in our simulations for the @xmath49-method + @xmath228 & level where observed lines transition from emission to absorption + @xmath229 & normalized radial wave function for level @xmath14 , @xmath35 + @xmath79 & ratio between the fine structure ( @xmath13-@xmath12 ) level population and the fine structure level population in lte + @xmath230 & integral of the radial matrix elements + @xmath33 & rydberg constant + @xmath231 & temperature of power law background spectrum at frequency @xmath232 + @xmath27 & electron temperature + @xmath233 & + @xmath125 & radiative recombination coefficient to a level @xmath14 + @xmath65 & radiative recombination coefficient to a level @xmath14 and sublevel @xmath35 + @xmath234 & dielectronic recombination rate + @xmath235 & correction factor for stimulated emission + @xmath236 & de - excitation rate for carbon ions in the @xmath13 core due to collisions with electrons + @xmath237 & de - excitation rate for carbon ions in the @xmath13 core due to collisions with hydrogen atoms + @xmath238 & energy difference between two levels + @xmath239 & @xmath240 , difference between the upper an lower principal quantum number + @xmath241 & correction factor to the planck function due to non - lte level population + @xmath120 & reduced mass + @xmath23 & frequency of a transition + @xmath232 & reference frequency for the power law background spectrum + @xmath22 & line profile + @xmath34 & statistical weight of level @xmath49 + @xmath37 & statistical weight of parent ion + @xmath242 & ionization potential of a level @xmath14 , divided by @xmath243 +
the strength ( or depth ) of an emission ( absorption ) line depends on the level population of atoms .
the line emission and absorption coefficients are given by ( e.g. @xcite ) : @xmath17 where @xmath18 is the planck constant , @xmath19 is the level population of a given upper level ( @xmath20 ) and @xmath21 is the level population of the lower level ( @xmath14 ) ; @xmath22 is the line profile , @xmath23 is the frequency of the transition and @xmath24 , @xmath25 are the einstein coefficients for spontaneous and stimulated emission ( absorption ) , respectively . following @xcite , we present the results of our modeling in terms of the departure coefficients ( @xmath42 ) and the correction factor for stimulated emission / absorption ( @xmath45 ) : @xmath244 @xmath245 unless otherwise stated the @xmath45 presented here correspond to @xmath246 , i.e. @xmath46 transitions .
when a cloud is located in front of a strong background source the integrated line to continuum ratio is proportional to @xmath147 @xcite .
we expand on the radiative transfer problem in paper ii . under thermodynamic equilibrium conditions ,
level populations are given by the saha - boltzmann equation ( e.g. @xcite ) : @xmath247 where @xmath28 is the electron density in the nebula , @xmath29 is the ion density , @xmath30 is the electron mass , @xmath31 is the boltzmann constant , @xmath33 is the rydberg constant @xmath34 is the statistical weight of the level @xmath14 and angular quantum momentum level @xmath35 [ @xmath36 , for hydrogen ] , @xmath37 is the statistical weight of the parent ion .
the factor @xmath248 is the thermal de broglie wavelength , @xmath249 , of the free electron @xmath250 $ ] . in general ,
lines are formed under non - lte conditions and , in order to properly model the line behavior , the level population equation must be solved .
we follow the methods described in @xcite and improved upon by @xcite as described in section [ section_method ] . here
, we give a detailed derivation of the theory and methods .
first , we solve the level population equation assuming statistical population of the angular momentum @xmath35-levels , i.e. : @xmath251 for all @xmath14 levels .
this assumption greatly simplifies the calculations but is only valid when @xmath35 changing transitions are faster than other processes , and , in general , this is not the case for low @xmath14 levels .
the level population equation under this assumption is ( e.g. @xcite ) : @xmath252 & = & \sum\limits_{n ' > n}{n_{n ' } a_{n'n}}+\sum\limits_{n ' \neq n}{n_{n'}(b_{n'n}i_\nu + c_{n'n } ) } \nonumber\\ & & + n_e n_{ion}(\alpha_n+ c_{in}).\end{aligned}\ ] ] the right- and left - hand side of equation [ eqn_levelpop1 ] describe how level @xmath14 is populated and depopulated , respectively .
we take into account spontaneous transitions from level @xmath14 to lower levels ( @xmath102 ) , stimulated emission and absorption ( @xmath253 , @xmath254 ) , collisional transitions ( @xmath255 ) , radiative recombination ( @xmath125 ) , collisional ionization ( @xmath256 ) and 3-body recombination ( @xmath257 ) .
equation [ eqn_levelpop1 ] can be written in terms of the departure coefficients ( @xmath42 ) : @xmath258 & = & \sum\limits_{n ' > n}{b_{n ' } \frac{\omega_{n'}}{\omega_n } e^{\delta \chi_{n'n } } a_{n'n}}\nonumber\\ + \sum\limits_{n ' \neq n}{b_{n ' } \frac{\omega_{n'}}{\omega_n } e^{\delta \chi_{n'n}}(b_{n'n}i_\nu + c_{n'n})}+\frac{n_e n_{ion}}{n_n(lte)}(\alpha_n+ c_{in}).\end{aligned}\ ] ] the previous equation can be written as a matrix equation of the form * r*@xmath259*b*=*s * by choosing the appropriate elements to form the matrices * r * and * s * ( e.g. @xcite ) : @xmath260 it is easy to solve for the @xmath42 values by using standard matrix inversion techniques .
we will refer to this approach of solving the level population equation as the @xmath14-method . at low @xmath14 levels ,
the quantum angular momentum distribution must be obtained , since the assumption that the angular momentum levels are in statistical equilibrium is no longer valid . moreover , as described in @xcite , dielectronic recombination is an important process for carbon ions at low temperatures and densities .
since the dielectronic recombination process depends on the quantum angular momentum distribution , we need to include the @xmath35 sublevel distribution for a given @xmath14 level .
the level population equation considering @xmath35-levels is : @xmath261 & = & \nonumber \\ \sum\limits_{n ' > n}\sum_{l'=l\pm1}{b_{n'l ' } \frac{\omega_{n'l'}}{\omega_{nl } } e^{\delta \chi_{n'n } } a_{n'l'nl}}+\sum\limits_{n ' \neq n}\sum_{l'=l\pm1}{b_{n'l ' } \frac{\omega_{n'l'}}{\omega_{nl } } e^{\delta \chi_{n'n}}(b_{n'l'nl}i_\nu + c_{n'l'nl } ) } \nonumber\\ + \sum_{l'=l\pm1 } b_{nl'}\left(\frac{\omega_{nl'}}{\omega_{nl } } \right)c_{nl'nl}+\frac{n_e n_{ion}}{n_{nl}(lte)}(\alpha_{nl}+ c_{i , nl}).\end{aligned}\ ] ] to solve for the @xmath35 level distribution at a given @xmath14 level we followed an iterative approach as described in @xcite .
we will refer to this approach of solving the level population equation as the @xmath49-method .
we start the computations by applying the @xmath14-method , i.e. assuming @xmath262 for all @xmath35 levels , thus obtaining @xmath263 values . for levels above a given @xmath98 value we expect the @xmath35-sublevels to be in statistical equilibrium . in this case ,
equation [ eqn_neqnl ] is valid and the @xmath40 values are equal to those obtained by the @xmath14-method . on the first iteration , we start solving equation [ eqn_llevelpop ] at @xmath264 and use the previously computed values ( @xmath265 ) for levels @xmath92 .
equation [ eqn_llevelpop ] is then a tri - diagonal matrix ( only elements with @xmath266 , enter in the equation ) and , by solving the system of equations , we obtain @xmath267 values .
the operation is repeated for all @xmath14 levels down to @xmath268 . in all our simulations
we assume @xmath269 since we are focused on studying carbon atoms whose ground level correspond to @xmath270 .
we repeat the operation by using the @xmath267 values instead of the @xmath263 values . @xcite
have proven that considering collisions from ( and to ) all @xmath20 levels guarantees a continuous distribution between both approaches at levels close to @xmath98 .
the final @xmath42 values are computed by taking the weighted sum of the @xmath40 values : @xmath271 details on the parameters used in this work are given in the text ( section [ section_nummethod ] ) .
in general , the radiative decay depends on the angular momentum quantum number of the electron at the level @xmath14 .
transitions from level @xmath272 are described by @xmath57 coefficients , in the dipole approximation @xcite : @xmath273 where @xmath119 is the bohr radius and @xmath229 is the normalized radial wave function solution to the schrdinger equation of the hydrogen atom @xcite .
the computation of the matrix elements is challenging ( see @xcite for details ) and we follow the recursion relations given by @xcite to calculate them up to @xmath274 . defining : @xmath275 where the first argument of @xmath230 corresponds to the lower state .
for a given @xmath20 level , @xcite give the following relations , with the starting values : @xmath276^{1/2 } \frac{(n - n')^{n - n'-2}}{(n+n')^{n+n'+2}}.\end{aligned}\ ] ] the recursion relations are : @xmath277 and : @xmath278 with : @xmath279
@xcite give a formula for computing the photoionization cross - section : @xmath280 to obtain the radial matrices elements , we use the same recursion formula as for the einstein a coefficients with the substitution : @xmath281 , with @xmath282 the imaginary number .
the @xmath213 coefficients are : @xmath283 and the initial values are : @xmath284^{1/2 } ( 4n')^{n'+2}e^{-2n'},\nonumber\\ r(n'-1,n')_{\kappa \neq 0}&= & \left[\frac{\prod_{s=1}^{n ' } ( 1+s^2\kappa^2)}{1-\mathrm{exp}(-2\pi/\kappa ) } \right]^{1/2 } \frac { \mathrm{exp}[2n'-(2/\kappa ) \mathrm{arctan}(n'\kappa)]}{\left(1+n'^2\kappa^2\right)^{n'+2 } } r(n'-1,n')_{\kappa=0}\nonumber.\end{aligned}\ ] ] we are interested in computing the recombination cross - section for an electron with energy @xmath285 recombining to a level @xmath49 . from milne relation
we obtain ( e.g. @xcite ) : @xmath286 expressed in terms of the radial matrices . here
, @xmath287 is the ionization energy of the level @xmath14 .
the final rate is obtained by integrating the cross - section over a maxwellian velocity distribution : @xmath288 we consider @xmath289 and @xmath290 is the function in the integral . to integrate the cross - section
, we followed an approach similar to @xcite .
we divide the integral in 30 segments starting at @xmath291 , and ending at @xmath292 .
each segment is integrated by using a 6-point gauss - legendre quadrature scheme .
this approach provides the value of the integral close to @xmath293 , therefore two correction factors must be applied : for the small values of @xmath294 we note that the integrand is almost constant and the value of the integral is then @xmath295 ; for large values of @xmath294 we use a 6-point gauss - legendre quadrature starting at @xmath296 and ending at @xmath297 .
as mentioned in section [ section_results ] we compare the sum over @xmath35 , of our radiative recombination rates with the formula of @xcite : @xmath298 with @xmath299 , and @xmath300 values for the @xmath301 are given by @xcite in two approximations for large and small argument , and tabulated values are also given for values in between the approximations . a first order expansion of the gaunt factor @xcite provides an accurate formula for the recombination coefficient : @xmath302
@xcite obtained the following semi - empirical formula for excitation by electrons .
the formula is given by : @xmath303,\end{aligned}\ ] ] with the coefficients defined as : @xmath304 @xmath305 \left ( 6 + 1.6 n s + \frac{0.3}{s^2}+0.8\sqrt{\frac{n^3}{s}}\left|s-0.6\right| \right)^{-1}.\\\end{aligned}\ ] ]
we use the formulation in the code of @xcite to obtain the values for the collisional ionization rates , the formulation is based on @xcite : @xmath306,\end{aligned}\ ] ] in units of @xmath307 . | in the first paper of this series , we study the level population problem of recombining carbon ions .
we focus our study on high quantum numbers anticipating observations of carbon radio recombination lines to be carried out by the low frequency array ( lofar ) .
we solve the level population equation including angular momentum levels with updated collision rates up to high principal quantum numbers .
we derive departure coefficients by solving the level population equation in the hydrogenic approximation and including low temperature dielectronic recombination effects .
our results in the hydrogenic approximation agree well with those of previous works .
when comparing our results including dielectronic recombination we find differences which we ascribe to updates in the atomic physics ( e.g. , collision rates ) and to the approximate solution method of the statistical equilibrium equations adopted in previous studies .
a comparison with observations is discussed in an accompanying article , as radiative transfer effects need to be considered . |
the recent discovery of superconductivity @xcite in na@[email protected]@xmath1o has aroused an active interest of solid state physicists , since this compound is regarded as a strongly correlated superconductor on a triangular lattice .
for such magnetically frustrated lattices , novel superconducting ( sc ) features are expected , as compared to the celebrated cuprates on bipartite lattices . in particular
, it is primarily important to determine the symmetry of its pairing potential .
so far , there have been a number of experimental efforts trying to fix the pairing symmetry , e.g. measurements of nuclear spin lattice relaxation rate @xmath18 @xcite , knight shift @xcite by nmr , and by @xmath19sr @xcite . at present , however , the results of these researches are not necessarily consistent to each another ; the pairing symmetry has not yet been established .
as for theory , although some studies regard multiband effects as important , @xcite many treat this issue with reduced single - band models on a triangular lattice .
the single - band studies are roughly classified into two categories , namely ones applying weak - coupling techniques to hubbard - type models , like perturbation expansions @xcite , a random - phase approximation @xcite and a fluctuation exchange approximation ( flex ) @xcite , and ones starting with @xmath4-@xmath5-type models in the strong - coupling limit.@xcite although the weak - coupling studies took mutually different approaches for different models , they arrived at a conclusion that @xmath13-wave spin - triplet symmetries are competitive with or even dominant over @xmath9-wave spin - singlet symmetries . on the other hand for @xmath4-@xmath5-type models
, it is unanimously shown that @xmath20+@xmath21-wave symmetries are widely stable against other symmetries as well as metallic states , using slave - boson mean - field approximations@xcite , which barely respect the local constraint of no double occupation , and a gutzwiller approximation@xcite , which takes account of the constraint on an average .
thus , the mean - field results of the strong - coupling theories disagree with the results of the weak - coupling ones , in contrast with the case of a square lattice , where the conclusion of the @xmath20 wave has been drawn in both theories . in this paper , we study the @xmath4-@xmath5 model on a triangular lattice , using a variational monte carlo ( vmc ) method , which exactly treats the local constraint of no double occupation .
our main purposes are ( i ) to determine the pairing symmetry of superconductivity within the @xmath4-@xmath5 model with this more accurate method , and ( ii ) to check the possibility of spin - triplet superconductivity in this strong - coupling model . besides the sc states , ( iii ) we consider the stability of nagaoka ferromagnetism@xcite and inhomogeneous phases .
based on these calculations , we argue the propriety of modeling for cobaltates .
the organization of this paper is as follows : in 2 , the formulation used is described .
sections 3 and 4 are assigned to the results and discussions on the stability of the singlet and triplet sc states , respectively . in 5 , we treat ferromagnetism and phase separation . in 6 , we recapitulate our results and address the relevance to cobaltates .
a part of the present results has been reported.@xcite
generally , the 3@xmath9 orbitals are split into @xmath22 and @xmath23 orbitals by an octahedral crystal field . in this material ,
the @xmath22 orbitals of co are further split into the levels of @xmath24 @xmath25 $ ] and @xmath26 @xmath27 $ ] @xcite , by a shift of oxygen vertical to the coo@xmath1 layers . due to an lda calculation for na@xmath28coo@xmath1@xcite , the bands near the fermi surface [ fig .
[ fig : band](b ) ] are involved in a co @xmath22 manifold [ fig .
[ fig : band](a ) ] .
since na@[email protected]@xmath1o has the same structure as na@xmath28coo@xmath1 , except for the intercalation of water molecules and the sodium concentration , one may assume the conduction bands for na@[email protected]@xmath1o are similarly formed from the @xmath22 manifold .
the band which forms the large fermi surface around the @xmath29 point ( fig .
[ fig : band](b ) ) has dominant @xmath24 character , whereas the band which forms the small pocket fermi surfaces on the @xmath29-k lines have mixed @xmath24 and @xmath30 character .
thus , we call the band indicated by dashed ( thick solid ) lines in fig .
[ fig : band](a ) the @xmath24 ( @xmath30 ) band .
many previous studies have assumed that na@[email protected]@xmath1o is basically described by single - band models of the @xmath24 or @xmath30 band , as mentioned .
coo@xmath1 , following an lda calculation@xcite .
lower panels : dispersion of single - band tight binding models on a triangular lattice for @xmath15 and @xmath31 .
( b ) hole fermi surfaces of the coo planes in na@xmath28coo@xmath1 , due to the same lda.,width=302,height=170 ] along this line , we consider a single - band @xmath4-@xmath5 model on an extended square lattice , @xmath32 where @xmath33 .
the sum of site pairs @xmath34 is restricted as follows : @xmath35 and @xmath36 for the nearest - neighbor pairs , @xmath37 and @xmath38 for the next - nearest - neighbor pairs only in one diagonal direction , and @xmath39 otherwise ( see fig .
[ fig : model ] ) . except for the case of @xmath15 and @xmath40 ( 4 ) ,
where the degeneracy at @xmath41 is serious , we put @xmath42 and @xmath43 .
here , @xmath6 indicates the hole concentration : @xmath44 with @xmath45 ( @xmath46 : electron number , @xmath47 : site number ) .
this parameter setting makes no difference with the isotropic triangular lattice .
let us make a remark on the sign of @xmath4 . in the lower panels of fig .
[ fig : band](a ) , the tight - binding bands of the present model , @xmath48 with @xmath42 are depicted for @xmath31 and @xmath15 .
compared with the lda band structure ( upper panel of fig .
[ fig : band](a ) ) , it seems appropriate that the band eq .
( [ eq : dispersion ] ) with @xmath15 [ @xmath31 ] should be connected to the @xmath24 ( dashed lines ) [ @xmath30 ( bold lines ) ] band . in na@xmath0coo@xmath1@xmath2@xmath3h@xmath1o ,
the electron density is larger than 1 , i.e. electron - doped case . to treat electron - doped cases in the framework of @xmath4-@xmath5-type models , namely in the electron density of @xmath50 without double occupation , we take advantage of a particle - hole transformation . thereby , an electron - doped case with @xmath51 , which corresponds to the @xmath24 band ( lower left panel of fig .
[ fig : band](a ) ) , is mapped to a hole - doped case with @xmath52 in @xmath4-@xmath5 model of eq .
( [ eq : model ] ) .
we use this sign convention throughout this paper .
in fact , the @xmath15 band , which corresponds to the @xmath30 band , has a van hove singularity at @xmath14 , as the @xmath30 band does near the k points .
so far , the majority of theoretical studies@xcite have considered that the large fermi surface around the @xmath29 point mainly contributes , whereas some papers have regarded the isolated pocket fermi surfaces or the van hove singularities as important for superconductivity .
@xcite in this paper , we consider both cases , @xmath15 and @xmath31 .
since the cobalt atoms in the present compound are connected through edge - shared oxygen cages , the effective value of @xmath7 is expected to be fairly small .
accordingly , we principally discuss the region of small @xmath7 in this paper .
we move on to the method used . to consider strongly correlated systems , more accurate treatment of the local constraint is necessary beyond the mean field level . for this purpose , a variational monte carlo ( vmc )
method @xcite is appropriate . in this method , variational expectation values of many - body wave functions
are numerically estimated with a monte carlo procedure , in which the local constraint , e.g. @xmath53 in eq .
( [ eq : model ] ) , is exactly satisfied .
actually , it has been shown that the vmc method is effective for a variety of analyses of the @xmath4-@xmath5-type models on square lattices without ( or with small ) frustration @xcite .
at least , we believe this method is reliable to discuss the mutual stability among various pairing symmetries . in this paper , we apply this method to the model eq .
( [ eq : model ] ) to study the stability of various sc states . for @xmath4-@xmath5-type models ,
a gutzwiller - type variational wave function is a good starting point,@xcite unlike hubbard - type models.@xcite for the normal state , we use the original gutzwiller wave function @xmath54 ( @xmath55 : fermi sea).@xcite a sc state with a fixed electron number@xcite is given as , @xmath56 with @xmath57 anisotropy of the pairing symmetry is introduced in @xmath58 .
we consider @xmath8 , extended-@xmath8 ( two kinds ) , @xmath9 and @xmath9+@xmath10 waves as candidates for singlet pairing , and @xmath11+@xmath12 and @xmath13 waves for triplet pairing .
they are summarized in table i. symmetries of @xmath8 , @xmath11+@xmath12 and @xmath9+@xmath10 waves have no line node in @xmath58 , whereas @xmath9 and @xmath13 waves have line nodes .
the time - reversal symmetry is broken for @xmath9+@xmath10 and @xmath11+@xmath12 waves . in the present calculation
, @xmath59 is the only variational parameter , which is connected to the sc gap , in particular for a high - doping region , but is not the gap itself in contrast with weak - coupling theories .
except for half filling , we may consider @xmath60 with finite @xmath59 is a sc state , while at half filling @xmath60 is an insulating rvb state.@xcite we fix @xmath19 at @xmath61 , the chemical potential for the noninteracting system , because @xmath19 dependence of @xmath62 is small , and @xmath19 does not seem an essential variable in a fixed-@xmath46 formulation .
.[tab : table1 ] @xmath58 for various pairing symmetries . [ cols="^,^",options="header " , ] since the number of variational parameters is at most one in this study , an usual vmc procedure @xcite sufficiently works . to obtain an accurate energy difference ( condensation energy ) , we have taken a sufficient number of samples to reduce the statistical errors , and kept the sampling interval long enough to ensure statistical independence of the samples . here ,
we collect samples as many as @xmath63-@xmath64 , which yield the error in energy of approximately @xmath65 .
we use the systems of @xmath66 ( @xmath67=4 - 18 ) with the periodic - antiperiodic boundary conditions , and the electron densities satisfying the closed shell condition .
in this section , we discuss the symmetries of singlet pairing , @xmath8 , extended-@xmath8 , @xmath9 and @xmath9+@xmath10 waves . to begin with , we consider @xmath8-type pairings .
figure [ fig : swave ] shows the hopping energy @xmath68 and exchange energy @xmath69 for the homogeneous @xmath8-wave symmetry , as a function of @xmath59 . for both @xmath15 and @xmath31 , each of @xmath70 and @xmath71 is a monotonically increasing function of @xmath59 , indicating the energy minimum is at @xmath72 for any value of @xmath73 ; the @xmath8-wave state is not stable against the normal state .
we also confirmed for various values of @xmath6 that this situation does not change .
this result is natural for the homogeneous @xmath8 wave , which favors on - site pairing . as for the two kinds of extended-@xmath8 waves , types ( 1 ) and ( 2 ) ,
although @xmath70 is a rapidly increasing function of @xmath59 , @xmath71 has a shallow minimum at a finite value of @xmath59 .
this means these symmetries become stable for finite values of @xmath73 , unlike the homogeneous @xmath8 wave .
we confirmed , however , that the energy minimum is at @xmath72 for a small value of @xmath7.@xcite we thus conclude that there is no possibility of the @xmath8 and extended-@xmath8 waves realized in the regime of our concern .
we will not refer to these states , henceforth .
-wave state as a function of @xmath74 for @xmath75 .
the two cases , @xmath31 and @xmath15 , are simultaneously shown .
we take the origin of each energy as the value for @xmath72 , namely the normal state , @xmath76 and @xmath77 ; similar metrics are used in some figures following .
the system used has @xmath78 sites .
, width=283,height=377 ] next , we turn to the @xmath9- and @xmath9+@xmath10-wave pairings . figures [ fig : didet ] and [ fig : didej ] show @xmath70 and @xmath71 , respectively , of the @xmath9+@xmath10 wave as a function of @xmath74 . in this case , @xmath70 is again a monotonically increasing function of @xmath59 , whereas @xmath71 conspicuously decreases and has a minimum at considerably large values of @xmath59 .
thereby , the @xmath9+@xmath10 state becomes stable even for considerably small value of @xmath7 ( e.g. @xmath79 for @xmath75 and @xmath31 ) .
similar behaviors of @xmath80 and @xmath81 can be seen for the @xmath9-wave state , as seen in figs .
[ fig : det ] and [ fig : dej ] . from these figures
, we find that the @xmath9+@xmath10 and @xmath9 waves have similar energies .
to see actual energy reduction , we plot in fig .
[ fig : didetot ] the total energy @xmath62 ( @xmath82 ) for @xmath83 and @xmath75 and 0.12 , as an example . for both @xmath9 and @xmath9+@xmath10 waves
, @xmath62 has a clear minimum at large values of @xmath59 for both signs of @xmath4 ; the sc states of @xmath9 and @xmath9+@xmath10 waves are appreciably stable near half filling . for @xmath84 ,
the minimum of the @xmath9+@xmath10 wave is slightly lower than that of the @xmath9 wave.@xcite + @xmath10-wave symmetry for several values of the hole density @xmath6 .
( a ) @xmath31 and ( b ) @xmath15 .
the system size is @xmath78 .
statistical errors are smaller than the symbol sizes . , width=283,height=377 ] + @xmath10-wave symmetry for several values of the hole density @xmath6 .
( a ) @xmath31 and ( b ) @xmath15 .
the inset in ( b ) is the magnification of the minimum area .
the system is the same as fig.[fig : didet ] . , width=283,height=377 ] -wave symmetry for several values of the hole density @xmath6 .
( a ) @xmath31 and ( b ) @xmath15 .
the system size is @xmath78 . , width=283,height=377 ] -wave symmetry for several values of the hole density @xmath6 .
( a ) @xmath31 and ( b ) @xmath15 .
the inset in ( b ) is the magnification of the minimum area .
the system is the same as fig.[fig : det ] . ,
width=283,height=377 ] - ( solid symbols ) and @xmath9+@xmath10-wave ( open symbols ) states for @xmath75 and 0.12 .
the system is of @xmath78 , and @xmath83 .
( a ) @xmath31 and ( b ) @xmath15 . ,
width=283,height=396 ] to consider the @xmath6 dependence of energy , we define a condensation energy , @xmath85 , as the difference of @xmath62 between the minimized superconducting state and the normal state , @xmath86 for @xmath87 , the superconducting state is stable against the normal state . plotted in fig .
[ fig : dec ] is @xmath85 for the @xmath9 and @xmath9+@xmath10 waves ; @xmath88 is a rapidly decreasing function of @xmath6 . near half filling , the @xmath9+@xmath10-wave state is more stable than the @xmath9-wave state for both signs of @xmath4 , probably because the symmetry of @xmath89 of the former is more favorable to the band symmetry , which is not far from isotropic .
the difference of @xmath85 , however , again becomes indistinguishable , as @xmath6 further increases , and @xmath88 itself becomes small .
note that @xmath85 of the @xmath9 and @xmath9+@xmath10 waves remains finite up to @xmath90 for @xmath31 , and @xmath91 for @xmath15 .
namely , the range of @xmath6 where the @xmath9+@xmath10-wave state is dominant is much wider for @xmath15 than for @xmath31 .
this behavior agrees with the gutzwiller approximation@xcite even quantitatively .
this difference is understood as follows : for @xmath15 , the density of states @xmath92 increases as @xmath6 increases , and has a van hove singularity at @xmath14 , while , for @xmath31 , @xmath92 is a monotonically decreasing function of @xmath6 until the band bottom .
this difference yields different behaviors particularly in @xmath81 , as seen in figs .
[ fig : didej ] and [ fig : dej ] . for @xmath31 ,
the minimum in @xmath81 smoothly increases as @xmath6 increases for both @xmath9 and @xmath9+@xmath10 waves . on the other hand , for @xmath15
, the minimum in @xmath81 once decreases as @xmath6 increase , and has the lowest value at @xmath93 for the @xmath9+@xmath10 wave [ fig .
[ fig : didej](b ) ] ; for the @xmath9 wave , the minimum in @xmath81 is stationary for @xmath94 [ fig . [
fig : dej](b ) ] . for both waves ,
the decrease in @xmath81 is still conspicuous for @xmath6 as large as 0.6 .
for this reason , the stability of the sc is enhanced particularly near the van hove singularity point ; we will return to this point in [ sec : triplet ] . - and @xmath9+@xmath10-wave states as a function of doping rate @xmath6 for @xmath83 .
( a ) @xmath15 and ( b ) @xmath31 . in ( b ) , the data for @xmath14 ( the van hove singularity point ) is obtained under a different condition ; for details , see [ sec : triplet ] .
note that the range of abscissa is twice larger for @xmath15 .
the data for @xmath95 , 12 and 18 are simultaneously plotted ; system - size dependence is not severe.,width=283,height=396 ] the above vmc results that the @xmath9+@xmath10 wave is dominant qualitatively support the conclusions derived from mean - field approximations@xcite and a gutzwiller approximation.@xcite for comparison , we plot in fig.[fig : comparison ] @xmath96 of the @xmath9+@xmath10-wave state by the present calculations and that of the gutzwiller approximation .
although @xmath96 by vmc is somewhat lower than that by gutzwiller approximation , the two curves exhibit similar doping dependence . of the @xmath9+@xmath10-wave pairing between the present vmc and a gutzwiller approximation@xcite for @xmath15 and @xmath83.,width=283,height=188 ]
ferromagnetic fluctuation , favored by triplet superconductivity , has been predicted by an lda calculation,@xcite and actually observed in an nmr-@xmath18 measurement for a superconducting sample.@xcite in addition , many theories in the weak - coupling region have made conclusions of probable triplet superconductivity . in view of these suggestions , we study the stability of triplet symmetries , @xmath11+@xmath12 and @xmath13 waves , in this section . for triplet symmetries , the same vmc procedure is carried out in the @xmath97 sector .
having calculated energies for various parameter values , it is found that @xmath81 is always a monotonically increasing function of @xmath59 .
this is because the exchange term @xmath98 with @xmath99 in eq .
( [ eq : model ] ) is obviously unfavorable to a triplet state .
moreover , @xmath70 is also a monotonically increasing functions of @xmath59 , except for the case of @xmath40 and @xmath15 .
namely , the triplet superconducting states are always unstable against the normal state , excluding this special condition . in the following , therefore
, we concentrate on this special case , @xmath15 and @xmath14 , at which the fermi energy is situated just at the van hove singularity point . at this point
, severe level degeneracy arises due to the divergence of @xmath100 . as a result , for finite systems , the electron densities where the closed - shell condition is satisfied disappear near @xmath14 .
for example , this range is @xmath101 for @xmath95 , as seen in fig.[fig : dec](b ) . to avoid this difficulty
, we slightly change the value of @xmath102 from 1 to e.g. 0.999999 , which lifts the degeneracy and make it possible to study @xmath14 without breaking the closed - shell condition .
@xcite + @xmath12 , @xmath13 , @xmath9+@xmath10 and @xmath9 waves for @xmath103 ( @xmath15 ) and @xmath14 ( @xmath104 ) . here
, we assume @xmath105 .
the inset in ( b ) shows the behavior of the @xmath9 and @xmath9+@xmath10 waves in a wider range .
, width=283,height=396 ] in fig .
[ fig : triplet](b ) , @xmath81 is shown for four symmetries . as mentioned above ,
@xmath106 works disadvantageously to the triplet pairing symmetries ; @xmath81 for the @xmath11+@xmath12 and @xmath13 waves is again a monotonically increasing function of @xmath59 , in contrast with the behavior of the singlet pairing symmetries .
thus , the triplet sc states are never stabilized when @xmath7 increases . in fig .
[ fig : triplet](a ) , @xmath80 is depicted for the same symmetries .
note that @xmath80 for the @xmath13 and @xmath11+@xmath12 waves ( solid symbols ) has a shallow minimum at a finite value of @xmath59 . from this result
, we find that the @xmath13-wave state becomes stable against the normal state ( @xmath72 ) for small values of @xmath7 ( @xmath107 ) .
compared with the singlet pairing symmetries , however , the decrease in @xmath80 for the triplet pairings is much smaller than that for the @xmath9+@xmath10 and @xmath9 waves .
moreover , the singlet sc states gain energies in @xmath81 for finite values of @xmath73 .
we add the data of the @xmath9 and @xmath9+@xmath10 waves at this density ( @xmath83 ) to fig .
[ fig : dec](b ) ; we find that the van hove singularity appreciably contributes to superconductivity .
nonetheless , we will see that even this @xmath9 wave is defeated by ferromagnetism for small @xmath7 . to summarize ,
triplet superconductivity is not realized in the triangular @xmath4-@xmath5 model .
in this section , we discuss ferromagnetic states and inhomogeneous phases , which are inseparable from the @xmath4-@xmath5 model .
let us begin with the nagaoka ferromagnetism.@xcite since nagaoka s theorem ( in the limit of @xmath7 , @xmath108 ) holds for @xmath15 in a triangular lattice , a finite range of ferromagnetism is expected near this limit .
actually , it was shown using a variation scheme for the hubbard model on a triangular lattice with @xmath15 that a nagaoka itinerant ferromagnetism is widely stable.@xcite for the @xmath4-@xmath5 model with @xmath15 , recent studies using a high - temperature expansion@xcite showed the nagaoka region to be @xmath109 and @xmath110 , which is quite wider than that for a square lattice ( possibly a partial ferromagnetism ) , namely , @xmath111 and @xmath112.@xcite with these in mind , we look at our vmc results in the following . in fig . [
fig : ferro ] , we show @xmath80 for the complete ferromagnetism ( nagaoka state ) , @xmath113 , and the @xmath9+@xmath10-wave sc state , @xmath114 . for @xmath31 , @xmath113 is always higher than @xmath114 ; thus the complete ferromagnetism never takes place , because @xmath115 for the nagaoka state , while @xmath116 further decreases due to the contribution of @xmath106 .
inversely , for @xmath15 , @xmath113 is appreciably smaller than @xmath114 for most values of @xmath6 , thereby there exist a region for small @xmath7 where the nagaoka state is more stable than the @xmath9+@xmath10-wave state . ) of the @xmath9+@xmath10-wave sc state and those of the complete ferromagnetic state ( @xmath113 ) , estimated analytically through a spinless fermion model .
( a ) @xmath31 , and ( b ) @xmath15 . , width=283,height=377 ]
comparing @xmath113 and @xmath116 for finite values of @xmath73 , we determined the phase boundary , as shown in fig .
[ fig : phased](b ) .
the range of ferromagnetism is very broad , namely @xmath117 and @xmath118 , which is quantitatively consistent with a recent result of high - temperature expansion,@xcite and is markedly wider than that for the square lattice estimated through the same vmc method.@xcite consequently , for @xmath15 , it is unlikely that superconductivity is realized in the region of small @xmath7 and intermediate @xmath6 , even if we consider that the ferromagnetic region somewhat shrinks by improving the trial function of superconducting states .
-@xmath7 space constructed from the present calculations , for ( a ) @xmath31 and ( b ) @xmath15 .
the arrows on the vertical axis indicate the critical values of the phase separation estimated by the way ( ii ) .
note that the system - size dependence is insignificant .
, width=283,height=415 ] now , we turn to the phase separation in the triangular lattice . since the exchange term in eq .
( [ eq : model ] ) works as a kind of attractive potential , the @xmath4-@xmath5 model necessarily brings about a phase separation for large values of @xmath7 . for the one - dimensional lattice
, the boundary between homogeneous and inhomogeneous phases extends in the direction of large @xmath73 as @xmath6 decreases @xcite , whereas for the square lattice , the boundary approaches the origin [ @xmath119 @xcite . here
, we estimate the boundary in two distinct ways .
( i ) we suppose that the inhomogeneous state phase separates into the two hole densities , @xmath120 ( half filling ) and @xmath121 ( empty ) , and the energy of this state is estimated as @xmath122 , where we adopt , as @xmath123 , a precise value obtained by a quantum monte carlo method for the heisenberg model,@xcite in which the antiferromagnetic ( af ) state with the @xmath124 structure is realized .
the boundary is determined by comparing @xmath125 and @xmath116 .
note that the reliability of this estimation increases as @xmath6 increases , but the boundary necessarily converges to @xmath126 as @xmath108 .
@xcite ( ii ) instead , we estimate the boundary near half filling , using the condition of intrinsic stability , @xmath127 , for the @xmath9+@xmath10-wave state . since we actually substitute finite differences for @xmath128 , the critical value of @xmath7 at half filling for @xmath129 is likely to be smaller
. the results of ( i ) and ( ii ) are summarized in fig .
[ fig : phased ] . for @xmath15 , the critical value @xmath130 is relatively small and barely depends on @xmath6 for large @xmath6 . for @xmath31
, @xmath130 is fairly large for large @xmath6 , and gradually decreases as @xmath6 decreases . for both signs of @xmath4
, @xmath130 abruptly decreases as @xmath6 approaches half filling ; this tendency is analogous to the square lattice case .
although the possibility of phase separation still remains for a realistic value of @xmath7 near half filling , this is not the case with intermediate and large values of @xmath6 .
in this paper , we have studied the single - band @xmath4-@xmath5 model on a triangular lattice , based on variational monte carlo ( vmc ) calculations , with na@[email protected]@xmath1o in mind .
the present study , which treats the local correlation precisely , is more reliable than the previous mean - field - type calculations .
we have compared the total energy for various pairing symmetries in a small-@xmath7 region , changing the doping rate @xmath6 and the sign of @xmath4 .
we consider @xmath8 , extended-@xmath8 , @xmath9 , and @xmath9+@xmath10 waves for singlet pairing , and @xmath11+@xmath12 and @xmath13 waves for triplet pairing .
furthermore , we have studied the stability of nagaoka ferromagnetism and inhomogeneous states .
main results are summarized as follows .
\(1 ) for @xmath31 , the @xmath9+@xmath10-wave state , which breaks time reversal symmetry , is dominant for @xmath131 ( @xmath132 ) , but the simple @xmath9-wave state has a quite close energy to that of the @xmath9+@xmath10 wave for a similar range of @xmath6 .
the stability of the superconducting state is rapidly lost , as the carrier density goes away from half filling . for @xmath84 ,
phase separation may take place .
see fig .
[ fig : phased](a ) .
\(2 ) for @xmath15 , in a wide parameter range of our concern ( @xmath133 , @xmath134 ) , the nagaoka ferromagnetic state defeats the @xmath9+@xmath10-wave state , which is the most stable among the superconducting states studied . see fig .
[ fig : phased](b ) .
\(3 ) triplet pairing is not realized in the @xmath4-@xmath5 model , although an @xmath13-wave state becomes slightly more stable than the normal state in the special case of @xmath14 and @xmath15 , where the fermi surface overlaps the van hove singularity points .
these results broadly support the previous mean - field - type studies,@xcite and is consistent with the high - temperature expansion studies.@xcite now , let us check these results for the triangular @xmath4-@xmath5 model in the light of a model of na@[email protected]@xmath1o . within the @xmath4-@xmath5 model with small @xmath7 , the range of steady superconductivity
is limited to @xmath131 with @xmath31 , and the pairing symmetry is the @xmath9+@xmath10 ( or possibly pure @xmath9 ) wave . in experiment , although the pairing symmetry of na@[email protected]@xmath1o has not yet established , it is likely that its gap has a line node,@xcite and does not break the time - reversal symmetry.@xcite furthermore , the carrier density at which superconductivity appears is limited to a narrow range near @xmath135.@xcite therefore , it is difficult to conclude that the @xmath4-@xmath5 model successfully describes the properties of na@[email protected]@xmath1o in the present experimental situation . provided that this discordance originates in the theoretical side , we discuss some related issues briefly in the remainder . as mentioned in 1 , the results of triplet , especially @xmath13-wave , pairing were obtained by applying weak - coupling approaches to hubbard - type models.@xcite among them , kuroki paid attention to the role of @xmath26 band@xcite , which broadly corresponds to the @xmath15 band .
they showed using a flex approximation that the @xmath13-wave state is the most stable .
thus , it is interesting to confirm whether the stabilization of the @xmath13 wave found for @xmath40 and @xmath15 in the @xmath4-@xmath5 model is connected to this weak - coupling @xmath13 wave in the hubbard model .
actually , the hubbard - type model became tractable by a vmc method due to a recent improvement of the variational wave functions.@xcite in this paper , we have used a single - band model as a first step to study the co - based oxide .
however , there exists a possibility that the multiband effect plays an essential role for superconductivity.@xcite actually , recent flex and perturbation calculations @xcite performed on multi - band hubbard models yielded results of dominant @xmath13 and @xmath11 waves . here , the @xmath26 band again makes a key contribution to superconductivity . to study this effect for large @xmath136
is also an interesting remaining issue . in the connection of triplet pairing
, we finally remark that the pairing symmetry of na@[email protected]@xmath1o can be determined by the phase sensitive experiments like tunneling effect and josephson effect through the andreev bound states .
@xcite in particular , to identify the triplet pairing , it is effective to use the anomalous proximity effect predicted very recently.@xcite the authors thank kazuhiko kuroki for useful discussions .
this work is partly supported by grant - in - aids from the ministry of education , culture , sports , science and technology , by the supercomputer center , issp , university of tokyo , and by naregi nanoscience project , ministry of education , culture , sports , science and technology , japan , which enables us to carry out the calculations on the computers at the research center for computational science , okazaki national research institutes . the coefficient of each electron configuration in eq.([eq : wavefunc ] ) can be represented in a determinantal form , so that the effective algorithm of ref .
@xcite ) works for such bcs - type functions .
see , j. p. bouchaud , a. georges and c. lhuillier : .
for example , for @xmath138 , the extended-@xmath8 wave ( 1 ) has a minimum at a finite value of @xmath59 for @xmath139 for @xmath31 , and for @xmath140 for @xmath15 . for the type ( 2 ) ,
the energy minimum is at @xmath72 for at least @xmath141 .
however , we leave it for a future study whether the extended-@xmath8 wave becomes stable for larger values of @xmath7 , as in the square lattice.@xcite at half filling , the pure @xmath9 wave has a lower value of @xmath62 than the @xmath9+@xmath10 wave , for finite systems .
checking the system - size dependence ( @xmath142 and 18 ) , however , we have found the values of the two symmetries seem degenerate in the @xmath129 limit . in reality
, an antiferromagnetism with @xmath124 structure , which is outside the scope of this study , is realized at @xmath120 .
we believe this nel structure immediately breaks as carriers are doped , due to the substantial frustration . within the present sc wave functions ,
the best value of @xmath71 ( @xmath9+@xmath10 wave , @xmath145 ) at half filling is @xmath146 , which is approximately 5% higher than a value obtained using a projector monte carlo method , @xmath147 .
@xcite thus , @xmath148 holds for @xmath108 , because , then , @xmath80 vanishes .
h. yokoyama , y. tanaka , m. ogata and h. tsuchiura , .
y. asano , y. tanaka , k. kuroki and s. kashiwaya : to appear in physica c ( 2004 ) . y. tanaka and s. kashiwaya , . s. kashiwaya and y. tanaka , .
y. tanaka , y. v. nazarov , and s. kashiwaya : ; | as a model of a novel superconductor na@xmath0coo@xmath1@xmath2@xmath3h@xmath1o , a single - band @xmath4-@xmath5 model on a triangular lattice is studied , using a variational monte carlo method .
we calculate the energies of various superconducting ( sc ) states , changing the doping rate @xmath6 and sign of @xmath4 for small @xmath7 .
symmetries of @xmath8 , @xmath9 , and @xmath9+@xmath10 ( @xmath11+@xmath12 and @xmath13 ) waves are taken up as candidates for singlet ( triplet ) pairing .
in addition , the possibility of nagaoka ferromagnetism and inhomogeneous phases is considered .
it is revealed that , among the sc states , the @xmath9+@xmath10 wave always has the lowest energy , which result supports previous mean - field studies . there is no possibility of triplet pairing , although the @xmath13-wave state becomes stable against a normal state in a special case ( @xmath14 and @xmath15 ) . for @xmath15 ,
the complete ferromagnetic state is dominant in a wide range of @xmath6 and @xmath7 , which covers the realistic parameter region of superconductivity . |
ngc1365 is a prominent barred spiral in the southern hemisphere .
its nuclear and disk emission - line gas has been investigated in numerous optical studies ( e.g. , burbidge & burbidge 1960 , osmer et al .
1974 , veron et al . 1980 , alloin et al .
1981 , edmunds & pagel 1982 , phillips et al . 1983 ,
jrster et al .
1984a , b , teuben et al .
1986 , edmunds et al .
1988 , schulz et al .
1994 , roy & walsh 1988 , 1997 ) .
the presence of an agn was first suggested by veron et al .
( 1980 ) who found broad emission - line h@xmath1 indicative of a seyfert-1.5 galaxy .
the fwhm of the broad component is 1900 km / s ( e.g. , schmitz 1996 ) .
surprisingly , just in the nucleus identified by edmunds & pagel ( 1982 ) , seyfert - typical narrow - line emission line ratios are missing , probably due to an interference with hii regions . to the se of the nucleus ,
[ oiii]@xmath2 enhancements and occasional line splitting apparently related to the agn are nevertheless present .
they trace a wide cone reminiscent of the supposed radiation cones in seyfert-2 objects ( jrster & lindblad 1989 , storchi - bergmann & bonatto 1991 , kristen et al .
outflow cone models were employed to fit the kinematics in this region ( phillips et al .
1983 , edmunds et al . 1988 ,
hjelm & lindblad 1996 ) . in the remaining circumnuclear emission - line region ,
hii region - like line ratios are common indicating widespread circumnuclear star formation .
two outstanding hot spots 7 sw of the center corroborate the starburst character .
sandqvist et al .
( 1995 ) mapped abundant molecular gas in the center which may provide the fuel for the birth of stars and which could have developed the molecular torus believed to be a prerequisite of seyfert unification ( antonucci 1993 ) .
summarizing , optical studies suggest that the central region of ngc1365 consists of an agn of apparent low luminosity surrounded by a region of enhanced star formation . however , the relationship between the stellar and nonthermal activity and the geometry of the nucleus need further scrutiny .
x - rays are an important probe of the central activity . in agns
, they are believed to arise in the innermost core .
interestingly , ngc1365 lies deep within the _ narrow - line _ galaxies in a plot of @xmath3 versus @xmath4 ( green et al .
1992 ) although its balmer lines have a conspicuous _ broad - line _ component signifying a class-1.5 seyfert . in an early study of pspc x - ray spectra by turner et al .
( 1993 ; tum93 hereafter ) a powerlaw fit was found to be unsatisfactory unless a line at 0.8 kev or a raymond - smith component was added .
recent detection of conspicuous fek line emission in data by iyomoto et al .
( 1997 ; i97 hereafter ) supports the view that the hard x - rays trace genuine nonthermal activity in ngc1365 .
in particular , the spectra revealed a striking similarity to those of ngc1068 , the prototype of a hidden class-1 seyfert ( antonucci & miller 1985 ) .
in the present work , we perform a detailed analysis of all x - ray observations of the core of ngc1365 and an extraordinary off - nuclear x - ray source which we term ngc1365x1 .
this investigation includes new pspc data and spectral models like a ` warm reflector ' , and the first high - spatial resolution x - ray study based on hri observations retrieved from the archive .
basic data of ngc1365 are adopted from the compilation in tab . 2 of schulz et al .
( 1994 ) : @xmath5 km / s implying a distance of 19.8 mpc ( linear scale 96 pc/ ) with @xmath6 km / s / mpc and the virgocentric model of kraan - korteweg ( 1986 ) .
recent hst borne cepheid data lead to an insignificantly smaller distance of @xmath7 mpc ( madore et al .
1996 , silbermann et al .
jrster & van moorsel ( 1995 ) found @xmath8 km / s and suggest a moderate revision of geometric data ( pa of line of nodes and value of inclination ) .
the paper is organized as follows : in sect.2 the x - ray spectrum of the nuclear source is analyzed while in sect.3 the high - spatial - resolution hri data revealing the core source and the variable enigmatic source ngc1365x1 are presented . in sect.4 ,
the properties of ngc1365x1 are investigated in more detail .
the nature of the sources is discussed in sect.5 which is followed by the concluding summary in sect.6 .
ngc1365 is serendipituously located in the field of view of a rosat ( trmper 1983 ) pspc ( pfeffermann et al .
1987 ) observation performed from jan .
30 feb . 2 , 1993 .
the total exposure time is 7.4 ksec .
the source is located at an off - axis angle of 30 . for further analysis ,
the source photons were extracted within a circular cell of radius 4 centered on the core of ngc1365 .
the background was determined in a source - free region near the target source and subtracted .
the data were corrected for vignetting and dead - time , using the exsas software package ( zimmermann et al .
the mean source countrate is @xmath9 cts / s . to carry out the spectral analysis source photons in the amplitude channels 11 - 240 were binned according to a constant signal / noise ratio of 5@xmath10 .
lllcl date & @xmath11 & @xmath12 & @xmath13 & obs .
label + pspc + jan 30 - feb 2 , 1993 & 7.4 & 0.102 & @xmath14 & pspc-1 + feb 5 - feb 10 & 7.7 & 0.096 & 0.005 & pspc-2 + hri + jul 20 - aug 4 , 1994 & 9.8 & 0.020 & 0.0026 & hri-1 + jul 4 - jul 5 , 1995 &
9.8 & 0.018 & @xmath150.001 & hri-2 + @xmath14 not resolved well enough in this off - axis observation ; thus , no countrate for ngc1365-x1 was determined a single powerlaw ( pl ) provides a marginally successful fit to the x - ray spectrum ( @xmath16 = 1.2 ) with a photon index @xmath17 , a 1-kev normalization of @xmath18 ph/@xmath19/s / kev , and cold absorption of @xmath20 @xmath21 which is larger than the galactic value in the direction of ngc1365 , @xmath22 @xmath21 ( dickey & lockman 1990 ) .
the fit becomes worse if we fix the cold absorption to the galactic value ( @xmath16 = 3.1 ) .
the pl fit leaves systematic residuals around 0.61 kev .
this new observation confirms the results of tum93 , who analyzed a different pspc observation and tentatively invoked a line of high equivalent width at 0.8 kev .
( re - analyzing this latter data set , using the same data reduction techniques as in the observation presented above and the same powerlaw model , we find both spectra to be consistent with each other . ) lllccc model & @xmath23 & @xmath24 & norm@xmath25 & @xmath26 & @xmath27 + & 10@xmath28 @xmath21 & & @xmath29 & kev & + pl & 0.8 & @xmath30 & 4.410@xmath31 & & 1.2 + rs & 0.135@xmath32 & & & 1.2 & 1.5 + rs@xmath33 & 0.22 & & & 1.1 & 0.9 + rs+pl-1.9 & 0.135@xmath32 & @xmath34 & 9.910@xmath35 & 0.95 & 0.9 + rs+pl-1.0 & 0.135@xmath32 & @xmath36 & 1.710@xmath31 & 0.72 & 0.5 + rs+rs & 0.135@xmath32 & & & 0.7/6.3 & 0.5 + @xmath37 fixed ; @xmath38 abundances of 0.2@xmath39cosmic ; @xmath29 pl normalization at 1 kev in ph/@xmath19/s / kev + we then applied models involving emission from a raymond - smith ( rs ) plasma .
if not noted otherwise , cosmic abundances ( allen 1973 ) were assumed throughout as further justified in sect .
5 . a one - component rs model yields an inferior fit , with @xmath16 = 1.5 .
the quality of the fit can be improved by allowing for low metal abundances . fixing them to 0.2 @xmath39 cosmic leads to @xmath26 = 1.1 kev , @xmath23 = @xmath40 @xmath21 , and @xmath16 = 0.9 . in the next step ,
two - component models were compared with the data .
both , a two - temperature rs model , and a rs+pl model provide excellent descriptions of the data . for the rs+rs model ( and @xmath23 fixed to the galactic value ) , we derive temperatures of 0.7 kev and 6.3 kev , with the hotter component dominating in flux ( @xmath16 = 0.5 ) . for the rs+pl description ( @xmath23 and photon index @xmath41 fixed ) we find a temperature of 0.95 kev ( @xmath16 = 0.9 ) . treating @xmath23 as additional free parameter results in a value slightly larger than the galactic value , but consistent with the latter within the errors .
the residuals for this model are shown in fig .
if we invoke a flatter underlying powerlaw spectrum , the temperature of the rs component slightly decreases .
e.g. , for @xmath42 ( about the value found in one of the fits in i97 ) and @xmath43 we get @xmath44 kev and @xmath45 .
the temperature rises somewhat , and the relative contribution of the rs component increases , if we allow for non - cosmic abundances ( e.g. , @xmath46 = 1.0 kev and @xmath16 = 0.9 for metal abundances of 0.3 @xmath39 cosmic ) .
finally , we checked whether the data can be described in terms of a lowly - ionized warm absorber , or a dusty warm absorber , or a warm reflector ( the models were calculated with ferland s ( 1993 ) code _ cloudy _ ; for assumptions and a detailed description see komossa & fink 1997a , b ) .
warm absorbers of low ionization or dusty ones imprint spectral signatures mainly at low energies beyond the sensitivity range ( e.g. , a carbon absorption edge at @xmath470.3 kev ) .
the underlying powerlaw index was fixed to the canonical seyfert value of @xmath48 ; solar abundances were adopted . whereas a warm absorber seen purely in absorption does not give a successful spectral fit , one in reflection provides an excellent description of the data .
the column density of the warm material is of order @xmath49 @xmath21 and the ionization parameter , where @xmath50 is the number rate of photons above 13.6 ev , @xmath51 is the distance nucleus warm absorber and @xmath52 the hydrogen density ] @xmath53 ( @xmath16 = 0.6 ;
the fit residuals are displayed in fig .
[ sedx ] ) .
the model is very similar to the scattering model that we applied to the x - ray spectrum of ngc6240 in komossa et al .
( 1998 ) .
the pspc x - ray lightcurve of ngc1365 was derived .
the x - ray emission is found to be constant . in the all - sky survey , ngc1365 was detected with a countrate of 0.09@xmath54 cts / s , within the errors consistent with the value of the pointed observations carried out later .
[ chi ] shows the pspc x - ray contours for the emission from the direction of ngc1365 .
four sources are detected , three of them very weak .
none of them coincides with any of the two supernovae reported in ngc1365 , sn 1957c and sn 1983v , but we note that ` companion 4 ' of tum93 ( their tab .
2a , 2b ) is at 16@xmath5520 distance to sn 1957c ( cf .
fig . 3 of komossa & schulz 1998 ) .
several source positions are marked in fig .
2 , including that of the enigmatic source ngc1365x1 .
whereas the x - ray emission maximum of the hri observations and a second pspc observation ( see below ) coincide well with the optical position of the nucleus , the maximum in the present pspc observation shows a slight off - set .
this can be traced back to the known boresight error of the telescope ( e.g. briel et al .
the region around ngc1365 was observed twice with the hri , from july 20 august 4 , 1994 ( labeled hri-1 hereafter ) and july 4 5 , 1995 ( hri-2 ) with an exposure time of 9.8 ksec each . both observations are centered on ngc1365 .
nine x - ray sources are detected with a likelihood @xmath56 within the field of view in the first observation , eight sources in the second . here , we focus on emission from the core of ngc1365 and the off - nuclear bright source rxj0333 3609 , hereafter referred to as ngc1365x1 . determining the background in source - free regions near the target sources
, we find background - corrected mean source countrates of 0.020@xmath57 cts / s and 0.018@xmath58 cts / s for ngc1365 for hri-1 and hri-2 , respectively .
ngc1365x1 is only detected in the first observation , with a countrate of 0.0026@xmath59 cts / s . in the second observation
, we estimate an upper limit on the countrate of @xmath60 @xmath15 0.001 cts / s .
we also re - analyzed another pspc observation of 7.7 ksec duration ( previously presented in tum93 ; these data are referred to as ` pspc-2 ' hereafter ) to check for the presence and brightness of ngc1365x1 .
we clearly detect this source ( note that it is none of the sources reported in tum93 ; their tab .
2b ) and derive a countrate of @xmath47 0.005 cts / s . in the hri observation
the two strongest central sources are clearly separated for the first time .
the stronger one coincides well with the nucleus of ngc1365 , the other source , ngc1365
x1 , lies to the south of the bar within or projected onto one of the ` side ' spiral arms ( fig . [ over ] ) .
the x - ray positions of the nuclear source ( j2000 ) are @xmath61 ( hri-1 ) and @xmath62 ( hri-2 ) , which compares to the position of the optical nucleus at @xmath63 taken from ned
. the coordinates of the x - ray emission maximum of ngc1365x1 are @xmath64 ( hri-1 ) . a contour plot of the x - ray emission traced by the hri is shown in fig .
[ conthri ] ( see also fig . [ over ] ) . due to its higher spatial resolution
the hri contours reveal well separated sources . because of the hri s lower sensitivity , some of the faint extranuclear pspc sources are undetected . `
companion 2 ' of tum93 is clearly resolved in two sources ( fig . [ over ] ) .
although the core source seems to be extended on weak emission levels , the presumably pointlike bl lac object in the field of view shows similar deviations from the instrumental psf . a deeper observation with , e.g.
, axaf would be needed to assess if part of the core emission is extended .
we note in passing that ngc1365x1 lies on a line connecting the nuclear source and the x - ray bright bl lac object ms03313 - 36 ; several further sources are ` aligned ' relative to the core source ( cf .
3 of komossa & schulz 1998 ; for a discussion of aligned x - ray sources in / around nearby galaxies see e.g. arp 1997 , radecke 1997 ) .
we produced long- and short - term x - ray lightcurves for both sources . whereas the emission from the core of ngc1365 and ngc1365x1 is constant within the errors on short timescales , that of ngc1365x1 is strongly variable on a longer timescale .
a clear detection during the first hri observation in 1994 was followed by a non - detection one year later implying a countrate drop by more than a factor 2 . for a comparison of the hri counts with the pspc and data
, we converted the counts from the latter instruments to equivalent hri counts , assuming a spectral shape as estimated in i97 ( @xmath65 and their galactic @xmath23 = 0.1510@xmath28 @xmath21 ; see also our sect .
the following countrates are obtained : @xmath66 = 0.0014 ( pspc-2 ; 2/1993 ) , 0.0026 ( hri-1 ; 7/94 ) , 0.0095 ( _ asca _ ; 1/95 ) , @xmath15 0.001 cts / s ( hri-2 ; 7/95 ) .
so far , the highest known state , corresponding to an isotropic luminosity of @xmath67 erg / s at the distance of ngc1365 , was reached during the observation ( i97 ) .
half a year later , during the hri-2 observation , the source had dropped by a factor @xmath68 10 as compared to the ` high state ' .
the x - ray lightcurve is displayed in fig .
[ light_offn ] .
given the extraordinary properties of ngc1365x1 , we made several checks to test the robustness of the results ( concerning amplitude of variability , derived luminosity , assumed spectrum ) : \(i ) first , we note , that measurements with the _ same _ instrument under similar conditions exhibit clear variability of the source : it is only detected in one out of the two hri observations which have the same exposure time and pointing direction .
ngc1365x1 is immediately visible in the first exposure , but absent in the second .
a similar argument holds for the two asca observations of i97 .
\(ii ) the total _ amplitude _ of variability depends on the inter - instrument countrate conversion ( from -sis to -hri ) . to test the reliability of the conversion
, we also converted the countrate of the central source ( which did not show variability during the or the observations ) to hri countrate and we find good agreement .
( explicitly , starting with cr = 0.023 cts / s ( sis ) and assuming constant spectral shape ( we used the rs model ) , we derive @xmath69 = 0.017@xmath58 , in agreement with the observed value of 0.019@xmath57 . ) \(iii ) the _ luminosity _ of the source depends on the assumed spectral shape and is particularly sensitive to the value of cold absorption .
i97 used the galactic @xmath23 only , which corresponds to a lower limit on source luminosity . to get a crude spectral shape in the band
, we used the on - axis pspc observation ( pspc-2 ) and extracted source photons in a circle around ngc1365x1 .
a potential problem is the closeness to the nuclear source , and the weakness of the target source .
clearly aware of being at the limit of the applicability of the chi - square fit procedure , we fit a powerlaw to the source spectrum .
in a first step , we fixed @xmath23 to the galactic value .
we derived a photon index of @xmath70 which is in good agreement with the value of @xmath71 .
we then repeated the spectral fit for a second background region ; the results are the same within the errors .
this insensitivity to the background chosen also implies that a potential contribution to the x - ray spectrum from other spiral arm sources in ngc1365 or weak extended emission from the fornax cluster is negligible .
we note that the observed ( isotropic ) high - state luminosity is a _ lower _ limit to the total luminosity ( except if dominated by beaming ) since the source may not have been observed exactly at maximum light , and the x - ray spectrum may extend into the euv and/or have a soft excess added .
\(iv ) finally , we recall that a foreground or background object can not be ruled out at present .
we consider this unlikely , though , due to the following reasons : ( a ) bright off - nuclear x - ray sources are being observed in more and more nearby galaxies ( the present case just seems to be an extreme one ) ; ( b ) foreground object ( e.g. a galactic x - ray binary ) : given the high galactic lattitude of ngc1365 ( @xmath72 ) a chance coincidence has a low probability ; ( c ) background quasar : located behind the dusty / gaseous spiral arm seen at the position of ngc1365x1 , the cold absorbing column in the direction of a background object should significantly exceed the galactic value .
however , the source is separately detected in the soft ( below 1 kev ) and the hard band .
we again used the crude pspc spectrum described above , fixed = 1.9 and treated @xmath23 as free parameter for an order - of - magnitude estimate of its value .
this yields @xmath73 @xmath21 and is consistent with the galactic value or only slightly larger .
again , we repeated the spectral fit for the second background region and get the same results within the errors .
cccc + + + + + +
a description of the soft x - ray spectrum in terms of a _ single _ pl ( sect.2.1 , see also fig .
1 ) leaves systematic residuals suggesting the presence of a thermal component .
unless metal abundances are strongly reduced ( a factor @xmath74 relative to cosmic ) a _ single _ rs model is unfeasible as well .
all evidence from hii regions in ngc1365 ( alloin et al .
1981 , roy & walsh 1988 , zaritsky et al . 1994
, roy & walsh 1997 ) leads to an o / h and n / h of above - solar or close - to - solar values in the center and a slight decrease outwards .
the shallowness of the gradient is explained by bar - induced _ strong mixing _ of the gas .
hence , strongly depleted metal abundances in the central x - ray gas are unexpected so that the successful metal - reduced single - rs fit is considered as an unlikely solution.solar abundances see also buote & fabian ( 1998 ) . ] in addition , the requirement of at least two components is in agreement with the data analysis by i97 . beyond the band they found a significant hard component which shows up as a very hot rs component ( @xmath75 kev ) in our rs+rs fit or is consistent with either a canonical pl ( @xmath76 ) plus a hot rs ( 1 kev ) or a flat ( @xmath42 ) pl and a ` moderate ' rs temperature ( 0.7 kev ) in our rs+pl fits .
a warm - absorber model seen _ purely _ in absorption did not fit the spectrum , either suggesting the lack of highly ionized gas directly along the l.o.s or a very high column density . aside from the amount of cold absorption , the two - component fits and the agn warm - scatterer model ( sect .
2.1 ) are strikingly similar to our results for the ultraluminous ir galaxy ngc6240 ( schulz et al .
1998 , komossa et al . 1998 ) in which we proposed an agn contribution in soft x rays . for ngc1365 ,
however , one of the arguments , the huge luminosity , can not be used against generation of the total x - ray flux in a starburst induced scenario .
our rs+pl@xmath48 model fit of the core source yields @xmath77 erg / s ( @xmath78 erg / s ) for @xmath23 free ( fixed to @xmath43 ) .
this is at least twenty ( more likely more than forty ) times less luminous than the central source of ngc6240 . in the following , to ease the luminosity demands , we use the solution for @xmath79 of which
the rs contribution is @xmath80 erg / s or 52% of the total .
the simplest idea one can have is to attribute the thermal rs component to a starburst source and the pl to the agn which means that they contribute in equal proportions in the band .
we take these numbers as a working hypothesis clearly having in mind that a precise decomposition is not possible at the present stage of measurement resolution and theory .
are we able to constrain the relative amounts of the agn and starburst contribution via multi - wavelength properties of ngc1365 ?
firstly , we formed ir gradients @xmath81 from the iras flux densities @xmath82 ( labeled by wavelength : @xmath83 jy , @xmath84 jy , @xmath85 jy ( @xmath86 ; retrieved from ned ) at @xmath87 and 100@xmath88 m and checked the location of ngc1365 in the @xmath89 diagram . samples of seyfert class-1 , class-2 and hii - type objects in this diagram are shown in miley et al.(1985 ) and mulchaey et al.(1994 ) . with @xmath90 and @xmath91
, ngc1365 lies in a region mainly occupied by hii - galaxies and seyfert-2 galaxies .
the slope indices signify dominating flux at smaller wavelengths ( ` warm ' dust components ) , warmer than typical seyfert-1s in the mid - ir ( 25,60 ) and warmer than the _ average _ of hii region type galaxies in the far ir ( 100,60 ) , making the fir cool dust component ( that dominates in normal spirals ) insignificant .
altogether this suggests that the ir emission comes from the central region where the agn and relatively young star formation components contribute a substantial portion of the observed ir emission in ngc1365 .
its more x - ray and ir - luminous sister ngc6240 contains a much stronger cool - dust fir - component ( @xmath92 ) but has nearly the same gradient in the mid - ir ( @xmath93 ) .
putting ngc1365 with @xmath94 into the scheme of heisler et al .
( 1997 ) shows a discrepancy .
this scheme assumes that objects with @xmath95 ( cold mid - ir , more edge - on view of torus ) should not even show a ` hidden ' blr .
but ngc1365 clearly exhibits broad balmer lines . converting the ratio into a power - law index @xmath96 ( 4 corresponds to @xmath97 ) and checking mulchaey s et al .
( 1994 ) plot of @xmath96 versus @xmath98 shows that at least six broad - line objects lie in heisler s et al .
( 1997 ) regime , where the edge - on view on the torus should hide the blr . a 100 wide [ oiii ] enhanced region with line - splitting to the se of the nucleus ( phillips et al .
1983 , jrster & lindblad 1989 ) was kinematically modelled as an outflow cone by hjelm & lindblad ( 1996 ) .
this could be a seyfert outflow , driven by a wind from the active core , or a starburst outflow , driven by a series of supernova explosions .
assuming that the outflow region evolved from a wind - driven supershell of swept - up ism we can use the maclow & mccray ( 1988 ) analytical model ( as e.g. applied by heckman et al .
( 1996 ) to arp220 and schulz et al .
( 1998 ) to ngc6240 ) for relating @xmath99 ( generated in the shell and the bubble interior to it ) to the mechanical input power @xmath100 .
if starburst driven , this can be derived from the sn rate by scaling the power of one sn per yr , @xmath101 erg / s .
kristen et al . ( 1997 ) estimate @xmath102 sn / yr for each of the two dominating brightest starburst knots 7 to the sw of the nucleus .
the nucleus has the same @xmath103-band brightness as one of these knots and may contain an equally strong mini burst plus the agn . as will be shown in sect .
5.1.4 , the optically detected agn could only provide a few percent of the ir .
assuming that starformation takes care of the rest we take the iras @xmath104 erg / s as an estimate for the bolometric luminosity of the starburst for which models from gehrz et al .
( 1983 ) predict a sn rate between 0.01 and 1 yr@xmath105 ( depending on the imf and the lower and upper mass limits ) and @xmath106 in the range @xmath107 erg / s of which kristen et al .
see only @xmath108 erg / s in the central region . hence ,
90% to 99% of the h@xmath1 emitters ionized by young stars are hidden .
could this hidden burst supply the x - rays in a wind driven shell ? from the velocity field in hjelm & lindblad ( 1996 ) we read off a radius @xmath109 kpc with a corresponding velocity @xmath110 km / s . with the thermal rs x - ray luminosity of @xmath80 erg / s
we get the following solution of the maclow - mccray equations given in scaled form as in schulz et al .
( 1998 ) : @xmath111 @xmath112 @xmath113 here , @xmath114 corresponds to @xmath115 erg / s or a sn rate of 0.015 yr@xmath105 which is consistent with the gehrz et al .
starburst models .
the expansion time @xmath116 corresponding to @xmath117 yrs and the density @xmath118 ( 1.2 @xmath119 ) are reasonable for the central region of this strong bar .
we conclude that this crude scenario can account for the observed thermal x - ray luminosity if more than 90% of the ionized gas emitting optical lines in the starburst region is obscured . due to less clear constraints
an analogous estimate for an agn wind is not attempted .
we now compare the data with the sample of hard x - ray detected agn from piccinotti et al .
concerning the two excellent agn tracers broad lines and hard x - rays this sample is homogeneous and comprises qsos , seyfert-1 galaxies and broad - line radio galaxies so that contributions by starburst components are minimized . for the piccinotti sample ,
ward et al .
( 1988 ) showed that the luminosities ( determined for @xmath120 km / s / mpc ) of hard x - rays , h@xmath1 and mid - ir - radiation from the flux parameter @xmath121 are well correlated ( their figs . 5 and 6 ) .
we represent their results by the relations @xmath122 the scatter in the ordinate is @xmath123 for both relations . for ngc1365 these quantities are observed to be ( cgs - quantities scaled to @xmath124=50 for the comparison ) : @xmath125 ( from i97 ) ; @xmath126 ( broad component measured by schulz et al .
1994 ) ; @xmath127 ( derived from iras data given by ned ) .
it turns out that these combinations of data do not fit the relations of the piccinotti sample for which we assume that it defines ` pure ' agn properties .
e.g. , if the hard x - rays come from the agn in ngc1365 we would expect @xmath128 which is only 0.43% of the observed 44.87 .
the remaining ir could either be attributed to star formation or the source for the hard 2 - 10 kev photons is for a large part obscured which would , however , require column densities exceeding @xmath129 @xmath21 . utilizing the first piccinotti - sample relation of eq . 4 ,
the x - rays predict @xmath130 , only 8.3% of the observed 40.82 .
the data were taken at different epochs , so variability can not be excluded to explain part of the discrepancy . however , the data for the piccinotti sample were collected non - simultaneously as well and their total scatter corresponds to only a factor four .
another possibiiity would be some arrangement of obscuring material in front of the x - ray source or / and the blr possibly combined with scattering elsewhere .
compton thick and balmer electron - scattering highly ionized thick columns might as well be envisioned in the extreme environment of the nucleus , but detailed models are beyond the scope of this work . taking h@xmath1 as representative for the agn , @xmath131 , i.e. ten times the observed hard x - ray luminosity , and @xmath132 , i.e. 2.5% of the observed mid - ir , would be predicted .
this type of solution would fit to our warm scatterer model ( last model in sect.2.2 ) which could explain both the hard pspc ( and _ asca _ ) pl - like component and the high equivalent width of the fek complex ( like in ngc6240 , see komossa et al .
1998 ) . a crude upper limit for the contribution from a stationary agn would be one that dominates the mid - infrared , leading to the prediction of @xmath133 = 44.22 and 43.05 for the hard x - ray and h@xmath1 luminosities , respectively . in this case , the corresponding observed ( see above ) hard x - rays and h@xmath1 luminosities are 0.06% of the predicted ones requiring esentially complete obscuration of the emission - line and hard x - ray sources .
such an extreme situation appears to be ruled out for ngc1365 because star formation is spatially extended over a diameter of @xmath134 and will provide an appreciable part of the ir .
ngc1365 does not fulfill expectations of the simplest version of the unified model in which only the torus blocks the light and detection of a blr would imply an unobscured view of the x - ray source as well .
complicated models appear to be necessary to let this agn be an _ intrinsically normal _ broad - line object .
other , currently vague , possibilities would be that the agn in ngc1365 as a low - luminosity object is qualitatively different from the piccinotti objects or that this object is an example for non - mainstream scenarios , like the starburst - agn models developed by terlevich et al .
however , such models have not yet been worked out to satisfactorily account for the hard x - ray data .
ngc1365x1 is a bright x - ray source .
the excellent spatial resolution of the rosat hri locates it on one of the spiral arms of ngc1365 .
the source is highly variable on the timescale of months . in sect .
4 we already gave cautious comments on the possibility of a foreground or background source and considered it unlikely . in the following discussion , we assume the source to be intrinsic to ngc1365 . in this case , ngc1365x1 is one of the brightest off - nuclear point sources known , with an isotropic luminosity of @xmath135 erg / s ( for @xmath136 mpc ) in the source s observed high - state .
off - nuclear x - ray sources , some of them variable , are frequently discovered in nearby galaxies ( e.g. fabbiano 1989 , vogler 1997 ; turner et al .
1993 , ehle et al .
1995 , brandt et al .
1996 , vogler et al .
1997 , supper et al .
1997 , immler et al .
their x - ray luminosities are typically of order 10@xmath137 erg / s up to several 10@xmath138 erg / s . among the most x - ray bright
observed sources in other galaxies are supernovae ( of type ii ; see e.g. the review by schlegel 1995 ) .
the reported maximum x - ray luminosities range between several 10@xmath138 erg / s ( e.g. sn 1993j , zimmermann et al .
1994b ; sn 1979c , immler et al .
1998 ) and 310@xmath139 erg / s ( sn 1986j , bregman & pildis 1992 ; all luminosities are as given by the authors ) . on long timescales ,
variability is detected in some sn ( cf .
schlegel 1995 ) . within our galaxy , one of the most luminous individual x - ray emitters is the superluminal - motion source grs 1915 + 105 with @xmath140 erg / s ( daily average during high - states ) and peak luminosities up to @xmath141 erg / s ( on a timescale of seconds ; greiner et al . 1998 ) .
ngc1365x1 is exceptionally luminous .
its temporal variability excludes an interpretation in terms of several spatially unresolved weak sources .
another definitely bright ( @xmath142 erg / s ) source was found by vogler et al .
( 1997 ) in an outer spiral arm of ngc 4559 .
they suggested an interpretation in terms of a ` buried supernova ' .
in contrast to ngc1365x1 , this source did not show any x - ray variability .
although a supernova in dense medium is an efficient way to reach high x - ray luminosities ( e.g. shull 1980 , wheeler et al .
1980 , terlevich et al . 1992 ) , the huge variability on the timescale of months we detect seems to favour an interpretation in terms of accretion onto a compact object . the amplitude and rapidity of variability share similarity with the off - nuclear source in ngc4945 reported by brandt et al .
( 1996 ) . however , the present case is much more powerful , by at least a factor of @xmath4716 in luminosity .
erg / s in the 0.82.4 kev band , ngc1365x1 shows @xmath143 erg / s in the same band . ] at present , the most likely interpretation seems to be an ultra - powerful x - ray binary , with either a highly super - eddington low - mass black hole or a very massive black hole . the hard powerlaw component derived from the and spectral fits
is known to be present in galactic x - ray binaries ( see tanaka 1997 for a recent review ) ; they have soft - excesses added in their high - states . in case
the accretion is not super - eddington , a rather high - mass black hole of @xmath47100200 m@xmath144 is inferred which may pose a challenge for stellar evolution models .
we studied the nuclear and circumnuclear x - ray emission of ngc1365 using pspc and hri observations .
the hri image shows compact emission from the nucleus , from the highly variable southwestern source , and resolves the source just outside the sw spiral arm into a double source .
the analysis of the spectral and temporal x - ray properties of the nucleus of ngc1365 and the enigmatic source rxj0333 - 3609 = ngc1365x1 is summarized as follows : + _ the core of ngc1365 . _ to describe the x - ray spectrum , we favour a two - component model consisting of about equally strong contributions from a raymond - smith ( rs ) plasma ( with @xmath47cosmic abundances ) and a powerlaw .
although a single rs model yields a successful spectral fit as well , it requires heavily depleted metal abundances which seems to be in contradiction to optical observations .
the powerlaw component signifies the presence of a hard component not common in simple starbursts ; this hard component as well as an fek line is also present in asca data of i97 strengthening the view that it arises in the agn .
the thermal rs part of the x - ray spectrum is found to be consistent with being produced in starburst - induced superwind shocks if the starburst dominates the mid to far ir and its h@xmath1 emitters are largely hidden . a detailed comparison of the x - ray properties with multi - wavelength observations of ngc1365 and a large sample of type - i agns was performed .
we find that , if the hard x rays ( 210 kev ) or the broad h@xmath1 are representative tracers of the active nucleus in ngc1365 , they can only account for a minute amount of the ir ; or , conversely , if the ir were due to dust heating by the agn , the latter has to be essentially completely hidden .
this and the large extent of the starformation region support the view that the ir is starburst dominated although emission - line h@xmath1 from the starburst has to be largely obscured as well .
a puzzle is the low x - ray luminosity of ngc1365 as compared to the strength of the blr balmer line component ; this poses a problem for the simplest version of the unified model .
in particular , a comparison with the piccinotti sample of type-1 agns strengthens the view that we see scattered x - rays in the 210 kev range .
this is in line with our alternative successful x - ray spectral model of a warm scatterer ; highly photoionized and high column density material close to the agn continuum source .
such a scenario , as alternative to a low - luminosity agn viewed directly , still requires a peculiar viewing geometry to simultaneously explain the observed strength of the broad lines .
+ _ ngc1365x1 .
_ attributed to ngc1365 , this source is among the most luminous ( @xmath145 several 10@xmath139 erg / s ) and most highly variable ( factor @xmath68 10 within 6 months ) off - nuclear x - ray sources known so far . at present
, the most likely interpretation seems to be an ultra - powerful x - ray binary . if this interpretation is confirmed , this or similar x - ray sources in nearby galaxies will be important probes for the extremes of stellar evolution .
further monitoring of ngc1365x1 , and derivation of improved spectral information will certainly be worthwhile .
st.k . acknowledges support from the verbundforschung under grant no .
50or93065 . it is a pleasure to thank per olof lindblad for helpful comments and suggestions .
the project is supported by the german bundesministerium fr bildung , wissenschaft , forschung und technologie ( bmbf / dlr ) and the max - planck - society .
we thank gary ferland for supplying _ cloudy _ , and andreas vogler for providing the software to plot the overlay contours of fig .
the optical image shown is based on photographic data of the national geographic society palomar observatory sky survey ( ngs - poss ) .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , caltech , under contract with the national aeronautics and space administration .
greiner j. , morgan e.h .
, remillard r.a . , 1998 , in _
frascati workshop on multifrequency behaviour of high energy cosmic sources _ , eds . f. giovanelli and l. sabau - graziati , mem .
1998 , in press ( http://www.obs.aip.de/ jcg / publis.html ) | we present an x - ray spectral and spatial analysis of the composite starburst / seyfert galaxy ngc1365 .
excellent fits of the pspc spectrum are obtained by combining a soft thermal component with a hard powerlaw .
the hard component may be either seen directly or can be explained by scattering of the agn powerlaw at circumnuclear warm high - column - density gas .
a compilation of the multi - wavelength properties of ngc1365 and comparison with hard x - ray selected agns shows that the hard component of ngc1365 is too faint compared to its broad balmer line components challenging simple unified models . according to analytical estimates ,
supernova driven outflow can fully account for the x - ray luminosity in the raymond - smith component if the observed ir emission is mainly provided by the central starburst .
we do not find obvious optical counterparts for three faint pspc sources south of the nucleus .
in particular , there is no coincidence with the two supernovae reported in ngc1365 . with the hri data ,
we have precisely located the extraordinary southwest x - ray source ngc1365x1 which falls on one of the subordinate spiral arms .
the source is found to be highly variable ( a factor @xmath0 10 ) on the timescale of months .
intrinsic to ngc1365 , its huge luminosity makes it exceptional among stellar x - ray sources . at present
, the most likely interpretation seems to be an ultra - powerful x - ray binary . |
convection - diffusion equations oftentimes arise in mathematical models for fluid dynamics , environmental modeling , petroleum reservoir simulation , and other applications . among them ,
the most challenging case for numerical simulation is the convection - dominated problems ( when diffusion effect is very small compared with the convection effect ) @xcite .
dominated convection phenomena could appear in many real world problems ; for example , convective heat transport with large pclet numbers @xcite , simulation of oil extraction from underground reservoirs @xcite , reactive solute transport in porous media @xcite , etc .
the solutions of these problems usually have sharp moving fronts and complex structures ; their nearly hyperbolic nature presents serious mathematical and numerical difficulties .
classical numerical methods , developed for diffusion - dominated problems , suffer from spurious oscillations for convection - dominated problems .
many innovative ideas , like upwinding , method of characteristics , and local discontinuous galerkin methods , have been introduced to handle these numerical difficulties efficiently ; see , for example , @xcite and references therein . for problems with nearly hyperbolic nature ,
it is nature to explore the idea of the so - called method of characteristics @xcite ; and , this idea has been combined with different spatial discretizations like finite difference ( fd ) , finite element ( fe ) , and finite volume ( fv ) methods . along this line of research ,
the semi - lagrangian method ( or , in the finite element context , the eulerian lagrangian method ) treats the convection and capacity terms together to carry out the temporal discretization in the lagrangian coordinate .
the eulerian lagrangian method ( elm ) gives rise to symmetric discrete linear systems , stabilizes the numerical approximation , and the corresponding diffusion problems are solved on a fixed mesh @xcite .
this method and its variants have been successfully applied not only on the linear convection - diffusion problem @xcite , but also the incompressible naiver - stokes equations and viscoelastic flow problems ; see , for example , @xcite . adaptive mesh refinement ( amr ) for partial differential equations ( pdes )
has been the object of intense study for more than three decades .
amr techniques have been proved to be successful to deal with multiscale phenomena and to reduce the computation work without losing accuracy when solution is not smooth . in general , the adaptive algorithm for static problems generates graded meshes and iterations in the following form : @xmath0 in the estimate procedure , we usually employ some computable local indicators to estimate the local error of the approximate solution we obtain from the solve procedure .
these indicators only depend on the datum and/or the approximate solution , and show in which part(s ) of the domain the local error is relatively too big or too small in the current mesh .
we then mark these subdomains and refine or coarsen them accordingly .
local error indicators determine whether the whole adaptive procedure is effective or not .
therefore , a reliable and efficient error indicator is the key to a successful amr method and a posteriori error analysis is often used to obtain such an error indicator in practice @xcite . in the context of finite element methods , the theory of a posteriori error analysis and adaptive algorithms for linear elliptic problem is now rather mature . convergence and optimality of adaptive methods for linear elliptic problems
have been proved as the outcome of a sequence of work @xcite ; see the recent review by nochetto , siebert , and veeser @xcite and references therein . on the other hand , for the nonlinear and time - dependent problems ,
the theory is still far from satisfactory .
a posteriori error analysis for nonlinear evolution problems is even more challenging .
adaptivity time - stepping is very important for time dependent problems because the practical problems sometimes have singularities or are multiscale in time .
uniform time step size can not capture these phenomena .
there are considerable amount of work in the literature devoted to the development of efficient adaptive algorithms for evolution problems .
a posteriori error estimators for linear parabolic problems are studied in @xcite and are also derived for nonlinear problems ; see @xcite for example .
there have been also some efforts for extending a posteriori error analysis for the time - dependent stokes as well as the navier - stokes equations @xcite . in particular ,
a posteriori error estimates for convection - diffusion problems have been discussed in @xcite .
it is nature to employ arm techniques for convection - dominated problems because of the complex structures of the solutions and evolution of these structures in time .
we also notice that spatial mesh adaptivity plays an important role in elm to reduce numerical oscillations and smearing effect when inexact numerical integrations are employed @xcite .
early adoption of adaptive characteristic methods has been seen since late 1980 s @xcite .
a posteriori error estimates for characteristic - galerkin method for convection - dominated problems have been proposed : demokowicz and oden @xcite considered the method of characteristics combined with petrov - galerkin finite element method for spatial discretization .
houston and sli @xcite give an a posteriori error estimator in @xmath1-norm for the linear time - dependent convection - diffusion problem using the duality argument . chen and ji @xcite give sharp a posteriori error estimations of elm in @xmath2-norm for linear and nonlinear convection - diffusion problems , respectively . a related @xmath3 a posteriori error bound can be found in chen , nochetto , and schmidt @xcite for the continuous casting problem ( convection - dominated and nonlinearly degenerated ) . in the previous error estimators mentioned above , the time residual , which measures the difference between the solutions of two successive time steps ,
is employed as a local indicator for temporal error . on the other hand ,
it is observed that elm is essentially transforming a convection - diffusion problem to a parabolic problem along the characteristics . in this paper
, we consider a posteriori error estimators for elm , with focus on temporal error estimators .
motivated by nochetto , savar , and verdi @xcite , in which adaptive time - stepping scheme for an abstract evolution equation @xmath4 in hilbert spaces is analyzed , we can obtain an a posteriori error bound for the temporal error along the characteristics .
combined with space adaptivity , adaptive method has been designed and implemented .
our numerical experiments in section [ sec : numer ] suggest that the new a posteriori error estimator is effective .
moreover , the numerical results also indicate that the proposed error estimators are very efficient and can take advantage of the fact that elm allows larger time stepsize .
the outline of this paper is as follows . in section [ sec :
elm ] , we introduce a model problem and its eulerian lagrangian discretization . in section [ sec : time ] , we discuss temporal a posteriori and a priori error analysis for the linear convection - diffusion problem . in section [ sec : full ] , we present a posteriori error estimation for the full - discretization . to simplify the discussion ,
we only consider a linear convection - diffusion model problem and restrict ourselves to the standard residual - type estimator for the spatial error , although the technique discussed in this paper can be potentially extended to other problems and spatial error estimators . in section [ sec
: numer ] , we consider the implementation of our adaptive algorithm and present some numerical experiments . throughout this paper , we will use the following notation .
the symbol @xmath5 denotes the space of all square integrable functions and its norm is denoted by @xmath6 .
let @xmath7 be the standard sobolev space of the scalar function whose weak derivatives up to order @xmath8 are square integrable , and , let @xmath9 and @xmath10 denote the standard sobolev norm and its corresponding seminorm on @xmath7 , respectively . furthermore ,
@xmath11 and @xmath12 denote the norm @xmath9 and the semi - norm @xmath13 restricted to the domain @xmath14 , respectively .
we also use the notation @xmath15 for the functions that belong to @xmath7 and their trace vanish on @xmath16 .
the dual space of @xmath17 is denoted by @xmath18 .
we use the notation @xmath19 to denote the existence of a generic constant @xmath20 , which depends only on @xmath21 , such that @xmath22 .
in this paper , we consider the following linear convection - diffusion model problem @xmath23\ ] ] with the initial condition @xmath24 and the boundary condition @xmath25.\ ] ] here @xmath26 ( @xmath27 ) is bounded polygonal domain .
we assume that @xmath28 is divergence free and vanishes on @xmath16 for @xmath29 $ ] .
we assume that @xmath30 and @xmath31 . in this section
, we briefly recall the construction of the eulerian lagrangian method . as usual , we introduce the characteristics ( particle trajectory ) @xmath32 , where @xmath33 is the _ lagrangian coordinate _
( original labeling ) at time @xmath34 of the particle .
it is also referred to the material coordinate .
and @xmath35 is the _ eulerian coordinate _ at the current time and referred to as reference coordinate .
then for the given velocity field @xmath36 , the characteristics is defind by the following ordinary differential equation : @xmath37 in the eulerian coordinates , be chain rule , the material derivative is defined as following @xmath38 and then we can rewrite as follow @xmath39 in order to avoid the deformation of the mesh , elm uses a fixed spatial mesh at each time - step and traces back along the characteristics .
the characteristics at each time interval @xmath40 $ ] ( @xmath41 ) has the same original labeling @xmath33 .
assume that @xmath42 and @xmath43.\ ] ] we discretize the material derivative in using the backward euler method as follows @xmath44 here , @xmath45 , @xmath46 , and @xmath47 .
it is easy to see that @xmath48 is a function of @xmath33 in the lagrangian coordinate .
we usually refer to as the temporal semi - discretization scheme . our a posteriori error estimate base on the assumption that the characteristics are solved exactly , which preserves the determinant of the jacobian of the flow .
in the numerical analysis , it is difficult to do that and it was apparent that a computational realization of preserving the determinant of the jacobian for the flow map to be one was crucial .
therefore , we discuss how to integrate the following ordinary differential equation for the computation of the characteristic feet .
the numerical scheme we discuss here has second order accuracy and preserves the determinant of the jacobian of the flow . @xmath49 where @xmath50 the equation ( [ eq : source ] ) is often called the source - free dynamical systems due to ( [ div : free ] ) .
for such a system , the solution @xmath51 is often called the flow map or phase flow and it has the following property . @xmath52 we shall begin with introducing some popular scheme to compute ( [ eq : source ] ) and showing that the scheme is indeed volume - preserving scheme for @xmath53 but not for @xmath54 .
we will then introduce some volume preserving scheme to solve ( [ eq : source ] ) for @xmath54 , which is due to feng and shang @xcite . in literatures , the following second order numerical scheme for solving ( [ eq : source ] )
is popular and it seems to first appear in @xcite .
first , we integrate the equation ( [ eq : source ] ) using the mid - point rule to obtain : @xmath55 where @xmath56 .
second , the right hand side is approximated by a second order accurate extrapolation .
namely , @xmath57 the following approximation shall also be used : @xmath58 for notational conveniences , let us denote @xmath59 and @xmath60 hence we have the following implicit approximations : @xmath61 to see that the mid - point rule ( [ mid : rule ] ) results in the volume - preserving scheme , let us take the derivative with respect to @xmath35 for both sides of ( [ mid : rule ] ) .
we then obtain the following : @xmath62 where @xmath63 and @xmath64 we solve ( [ eq : mid ] ) for @xmath65 and obtain that @xmath66 under the assumption that some appropriate finite element space for the velocity field is used so that the divergence free condition of the velocity field is imposed in the discrete sense , we have @xmath67 from ( [ eq : tr ] ) , we conclude that @xmath68 identically .
the main reason why such an algorithm is popular seems that it has the volume preserving property . on the other hand , it is easy to see that the algorithm may not result in the volume - preserving scheme for @xmath54 .
note that for @xmath54 , under the assumption that @xmath69 , for @xmath70 given as follows , @xmath71 we have that @xmath72 our purpose here is that by reviewing the volume - preserving scheme in three dimension developed by feng and shang in @xcite , we wish to make sure such a volume preserving scheme can be devised in three dimension and confirm our numerical scheme can be implemented .
as far as the author is concerned , such a special algorithmic detail has not been implemented in the context of the semi - lagrangian scheme .
the basic idea of constructing the volume preserving scheme for @xmath73 is based upon the following observation .
following the idea of h. weyl , we have : @xmath74 where @xmath75 the actual expressions for @xmath76 , @xmath77 , @xmath78 and @xmath79 as follows : @xmath80 from this , it is easy to see that ( [ hw : decomp ] ) holds and by construction @xmath81 and @xmath82 given as follows are divergence free : @xmath83 let us now denote @xmath84 by the volume preserving scheme for @xmath85 with @xmath86 , then the following composition is trivially volume preserving : @xmath87 moreover , assuming @xmath88 is of second order accurate , it is easy to see that the composition ( [ com : vol ] ) is of second order .
the above idea on the composition is from feng and shang , @xcite . discretizing with suitable spatial discretizations , we obtain fully discrete numerical schemes . in this paper , we focus on the finite element method .
first , we define the weak forms of and as usual .
we define the bilinear form @xmath89 as follow @xmath90 we denote the potential by @xmath91 and its frechet derivative by @xmath92 , i.e. , @xmath93 it is easy to see that @xmath94 is convex and @xmath92 satisfies the following inequality @xmath95 in fact , it is well - known that we have the following identity , @xmath96 where @xmath97 . furthermore , by taking the test function as @xmath98 in and applying , we obtain that @xmath99 the weak forms of ( [ eqn : sl_model ] ) and ( [ eqn : td_sl_model ] ) can be written as @xmath100 and @xmath101 respectively . on a shape - regular triangulation @xmath102 of @xmath21 , we introduce the piecewise continuous linear finite element space @xmath103 such that @xmath104 let @xmath105 and @xmath106 denote the mesh and the finite element space at time @xmath107 respectively .
then the fully discrete scheme can be written as : suppose @xmath108 is known , find @xmath109 such that @xmath110 where @xmath111 is some suitable approximation of @xmath112 .
in this section , we focus on the a posteriori error estimation for temporal semi - discretization ( [ eqn : td_sl_model ] ) . for the sake of simplicity , we assume that @xmath113 in this section .
different from the standard time error indicators for elm , which measure the solution difference along the time direction , our new error indicator measures the difference along the characteristics .
we will establish a posteriori error estimation based on the new time error indicator and show its efficiency .
we will also show an optimal priori error bound as a byproduct . here
optimality does not only mean the optimal convergence rate which can also be achieved by classic error analysis , but also mean the optimal regularity requirement which have not been proved by standard a priori error analysis .
we first show that the solution of satisfies an energy identity along characteristics .
in fact , by taking inner product with @xmath114 on both side of and applying the following identity @xmath115 we immediately obtain that @xmath116 we can see that the convection diffusion equation preserve the total energy along characteristics in continuous level . on the other hand , by choosing the test function @xmath117 in and employing , we have an discrete energy inequality : for any integer @xmath118 , @xmath119 we note that the equality holds in if there is no temporal discretization error .
this motivates the following definition : @xmath120 and we can view the computable quantity @xmath121 as a measure of the deviation of numerical solution from satisfying the energy conservation .
we can use @xmath122 as a time error indicator for adaptive time stepping in elm . in order to give an a posteriori error bound for our new time error indicator
, we construct the following linear interpolation @xmath123 where @xmath124 since the original labeling @xmath33 of the characteristics does not depend on time , we have @xmath125.\ ] ] substituting back to and then applying , we have @xmath126 by adding and subtracting @xmath127 , for any @xmath128 , we have @xmath129 where @xmath130 is the residual ( which does not depend on the choice of test function ) : @xmath131 we notice that @xmath94 is a convexity functional , i.e. , @xmath132.\ ] ] hence , @xmath130 can be bounded by @xmath133 by choosing @xmath134 in , and @xmath135 in , adding these two equations , and using , we end up with the following inequality @xmath136.\ ] ] then the following upper bound of the new time error indicator holds : [ thm : td_upper ] let @xmath137 be the exact solution of and @xmath138 be the time semi - discrete solution in .
assume that @xmath139 is defined as .
for any integer @xmath140 the folltowing upper bound holds : @xmath141 integration in time , we can obtain @xmath142 since @xmath28 is divergence free , which implies @xmath143 , we have , after changing of variables , that @xmath144 then follows directly from summing up above inequality from @xmath145 to @xmath146 .
@xmath147 traditional a priori error analysis for elm treats the temporal semi - discretization as a finite difference method and derive the error estimation based on the taylor expansion ( see @xcite ) . as a result
, we obtain an optimal convergence rate but the regularity requirement is suboptimal . taking advantage of the posteriori error estimation of new time error indicator @xmath148 , we can derive an optimal order priori error estimation with minimal regularity requirement on the datum and the solution . from the definition of the new a posteriori error estimator , and , we have @xmath149 by the definition of @xmath150 and , we have @xmath151 substitute back into , we have @xmath152 set @xmath153 and use the elementary inequality @xmath154 , we obtain @xmath155 now using the assumption that @xmath28 is divergence free and summing up from @xmath145 to @xmath146 , we obtain the following optimal a priori error estimation with minimal regularity requirement : let @xmath137 be the solution of with initial value @xmath156 , and @xmath157 is the time semi - discrete numerical solution of the temporal semi - discretization . for any integer @xmath158 , we have the following priori error estimate @xmath159 only on initial guess .
in this section , a posteriori error estimation for the fully - discrete scheme is considered .
the characteristic feet @xmath160 is defined by and is computed exactly . besides the new time error indicator , residual type error estimator is used as a spatial error indicator .
let @xmath161 be the _ interior residual _
, i.e. , @xmath162 and @xmath163 be the _ jump residual _ also defined , i.e. , @xmath164 where @xmath165 is the edge / face shared by @xmath166 and @xmath167 , @xmath168 is the unit normal vector of @xmath169 from @xmath170 to @xmath171 . as the analysis of time semi - discretization
, we define the linear interpolation @xmath172 of @xmath173 as following @xmath174 similar with , we have @xmath175 now , the following a posteriori error estimation holds [ thm : fd_upper ] let @xmath137 and @xmath176 be the solutions of and respectively .
for any integer @xmath158 , there exists a constant @xmath177 which does not depend on the mesh size and the time step size , such that the following a posteriori error bound @xmath178 holds , where the temporal error indicator @xmath139 and the spatial error indicator @xmath179 are defined as @xmath180 and @xmath181 for any @xmath182 and @xmath183 , we have @xmath184 let @xmath185 in and @xmath186 in , we have @xmath187 inserting @xmath188 , we have @xmath189 let @xmath190 and @xmath191 in , we have @xmath192 then let @xmath193 and @xmath194 where @xmath195 is the so called clement interpolation operator
. moreover , let @xmath196 in and add to above inequality , we can obtain that @xmath197 then by the definitions , and , we have @xmath198\mathrm{d}s \\ & + ( f - f_h^n , u- u_h )
\\ \le & ( t - t_{n } ) \xi_{n } + c \eta_{n}^{1/2 } |\!|\!| u - u_{h}^{n } |\!|\!|
+ |\!| f - f_h^n |\!|_{-a } |\!|\!| u- u_h |\!|\!|\\ \le & ( t - t_{n } ) \xi_{n } + c \eta_{n } + \frac{1}{4 } |\!|\!| u - u_{h}^{n } |\!|\!|^{2 } + \frac{1}{2}|\!|f - f^n_h|\!|^2_{-a } + \frac{1}{2 } |\!|\!|u - u_h|\!|\!|^2.\end{aligned}\ ] ] here in last two inequalities , we use the interpolation property of the clement interpolation , the cauchy schwartz inequality , and the young s inequality .
the norm @xmath199 is defined by @xmath200 .
now , we have @xmath201 for any @xmath202 $ ] , by integrating in time from @xmath203 to @xmath204 , we have @xmath205 for @xmath206 , by integrating in time from @xmath207 to @xmath208 , we have @xmath209 using the assumption that @xmath28 is divergence free , sum the above two inequalities from @xmath145 to @xmath146 , we have @xmath210 note that @xmath211 is arbitrary , the above inequality implies the upper bound .@xmath147
in this section , we first introduce the adaptive algorithm we use in numerical experiments and then present some numerical results .
since the main purpose of this paper is analyze the proposed time error estimator , we start with the algorithm for time step size control . let @xmath212 be the total tolerance allowed for the error introduced by the time discretization , according to theorem [ thm : fd_upper ] , that is @xmath213 therefore , a natural way to satisfy is to adjust the time step size such that @xmath214 based on this , now we propose the following algorithm to adjust the time step size and control the error under given tolerance @xmath212 .
[ alm : time_adaptive ] given tolerance @xmath212 , @xmath215 , @xmath216 , @xmath217 , and the initial time stepsize @xmath218 . 1
. set @xmath219 2 .
set @xmath220 3 .
solve the time semi - discretization problem using time step size @xmath221 4 .
compute the time error indicator 5 .
check the error : if is not satisfied , @xmath222 and goto ( 2 ) 6 .
if @xmath223 and @xmath224 , @xmath225 7 .
let @xmath226 and goto ( 2 ) the above algorithm is for adjusting the time step size and controlling the error introduce by time discretiziation .
when we consider the fully discretization , we need to take mesh adaption into account . according to theorem [ thm : fd_upper ] ,
the space error indicator gives us the elements where the local error is relatively large and should be refined .
let @xmath227 be the tolerance allowed for the error introduced by the spatial discretization , then the usual stopping criterion for mesh refinement is : @xmath228 as usual , in order to achieve equal distribution of error , can be replaced by @xmath229 the main difficulty for time - dependent problems is the mesh coarsening .
we need a coarsening error indicator to guide the mesh coarsening procedure . here
, we employ the same coarsening error indicator as the one proposed in @xcite .
let @xmath230 be a coarsening mesh from mesh @xmath105 , and @xmath231 and @xmath106 be the corresponding finite element spaces .
it is nature that @xmath232 .
furthermore , let @xmath233 and @xmath234 be the numerical solution at time @xmath107 satisfy that @xmath235 subtracting these two equalities and taking @xmath236 , we have @xmath237 using the elementary equality @xmath238 for the two terms on the left hand side of the above equality respectively , we can obtain the following galerkin orthogonal relation @xmath239 then we have @xmath240 together with the triangular inequality @xmath241 and theorem [ thm : fd_upper ] , this suggests us the following coarsening error indicator , @xmath242 and the stopping criterion for coarsening @xmath243 this indicator does not depend on the solution @xmath244 , and allows us to coarsen only once after mesh refinement , which avoid the undesirable case that the elements which were marked for coarsening must be refined again to reduce the total error .
there are several different ways to couple time step size control and mesh adaption .
a simple way is to include space adaptivity in every iteration of algorithm [ alm : time_adaptive ] . but
this may result in solving a large number of discrete systems and adaptivity may suffer due to the computational overhead .
here we apply a mild way to couple the time and space adaptivity which is a modification of the adaptive procedure in @xcite . at each time step , we first adjust the time step size with the old mesh , and then adapt the mesh , and followed by checking the time error again .
this leads to the following adaptive algorithm for on single time step .
given the tolerance @xmath212 , @xmath245 and @xmath246 , parameters @xmath215 , @xmath247 and @xmath248 .
let @xmath249 and the mesh @xmath250 have been obtained at the previous time step @xmath207 . 1 .
* set @xmath251 , @xmath252 and @xmath253 * solve the fully discrete problem on @xmath105 using time step size @xmath254 * compute the error indicators 2 .
while is not satisfied , * set @xmath222 and @xmath253 * solve the fully discretize problem on @xmath105 using time step size @xmath254 * compute the error indicators + end while 3 .
while is not satisfied , * refine the mesh and produce a refined @xmath105 * solve the fully discretize problem on @xmath105 using time step size @xmath254 * compute the error indicators * while is not satisfied , * * set @xmath222 and @xmath253 * * solve the fully discretize problem on @xmath105 using time step size @xmath254 * * compute the error indicators + end while + end while 4 .
* coarsen the mesh to produce a new mesh @xmath105 according to * solve the fully discretize problem on @xmath105 using time step size @xmath254 5 . if @xmath223 and @xmath224 , then * @xmath225 + end if now we will present some numerical results to demonstrate the efficiency and robustness of our new adaptive method . in our numerical experiments
, since we use backward euler to discretize the material derivative , we choose @xmath255 , @xmath256 and @xmath257 .
[ exp : peak-1d ] given the initial condition @xmath258 and @xmath259 , the exact solution for the model problem ( if @xmath28 is constant ) is given by @xmath260 where @xmath261 is a parameter measures the width of the support of the solution . the computational domain is @xmath262 $ ] , and we give the dirichlet boundary condition @xmath263 .
[ exp : shock-1d ] given the initial condition @xmath264 and @xmath259 , the exact solution for the model problem ( if @xmath28 is constant ) is given by @xmath265 where @xmath266 is the so - called complementary error function .
the computational domain is @xmath267 $ ] , and we give the dirichlet boundary condition @xmath268 . .
left : time step size comparison ; right : numerical solution comparison , title="fig:",scaledwidth=49.0%,height=240 ] . left : time step size comparison ; right : numerical solution comparison , title="fig:",scaledwidth=49.0%,height=240 ] ( @xmath269 ) . left : time step size comparison ; right : numerical solution comparison , title="fig:",scaledwidth=49.0%,height=240 ] ( @xmath269 ) . left : time step size comparison
; right : numerical solution comparison , title="fig:",scaledwidth=49.0%,height=240 ] figure [ fig : peak-1d ] and [ fig : shock-1d ] shows the numerical results for example [ exp : peak-1d ] and [ exp : shock-1d ] respectively .
we can easily see that the new time error indicator allows larger time step size , but the numerical solution still maintains the accuracy .
this means the the new time error indicator is reliable and effective .
[ exp : peak-2d ] this example is the benchmark gaussian - cone problem ( cf .
@xcite ) . given the velocity field @xmath270 and the initial condition @xmath271/(2\lambda^{2})\right\},\ ] ] and @xmath259 .
the exact solution is @xmath272/(2\lambda^{2}+4\varepsilon t)\},\ ] ] where @xmath273 , @xmath274 , @xmath275 , and @xmath276 .
the computational domain is @xmath277\times[0,1]$ ] in @xmath278 .
the following two tests are generalizations of the one - dimensional problem , problem [ exp : shock-1d ] .
[ exp : shock1 - 2d ] given the velocity field @xmath279 , and the initial condition @xmath280 and @xmath259 . then the exact solution is @xmath281 the computational domain is @xmath277\times[0,1]$ ] in @xmath278 . [ exp : shock2 - 2d ] given the velocity field @xmath282 , and the initial condition as @xmath283 and @xmath259 .
the exact solution is @xmath284 and the computational domain is @xmath277\times[0,1]$ ] in @xmath278 .
( @xmath269 ) . left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=33.0%,height=192 ] ( @xmath269 ) .
left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=33.0%,height=192 ] ( @xmath269 ) . left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=32.0%,height=192 ] ( @xmath269 ) . left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=33.0%,height=192 ] ( @xmath269 ) . left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=33.0%,height=192 ] ( @xmath269 ) . left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=32.0%,height=192 ] ( @xmath269 ) . left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=33.0%,height=192 ] ( @xmath269 ) . left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=33.0%,height=192 ] ( @xmath269 ) . left : time step size comparison ; middle : numerical solution ; right : adaptive mesh.,title="fig:",scaledwidth=32.0%,height=192 ] figure [ fig : peak-2d ] , [ fig : shock1 - 2d ] and [ fig : shock2 - 2d ] show the numerical results for example [ exp : peak-2d ] , [ exp : shock1 - 2d ] and [ exp : shock2 - 2d ] respectively .
as shown in the pictures , the new time error indicator allows larger time step size and the space error indicator captures the singularity .
overall , the adaptive finite element method based on our error estimate is effective and reliable for convection dominated diffusion problems .
in this paper , we discuss the adaptive elm for linear convection - diffusion equations .
we * derive a new temporal error indicator along the characteristics .
we are able to show optimal convergence rate of temporal semi - discretization with minimal regularity of the exact solution . *
combine the new temporal error indicator with residual - type spatial error indicator and obtain a posteriori error estimation for the fully discretization elm . * design efficient adaptive algorithm based on the a posteriori error estimators .
numerical results shows robustness of the new adaptive method and allows larger time steps compared with standard temporal error indicator for elm in general . for the future work
, we are working on generalize the a posteriori error estimation for nonlinear convection dominated problems , where the characteristics has to be solved approximately . in this case , ode solvers that preserves the determinant of the jacobian of the flow will be important .
meanwhile , we will also generalize the algorithm to navier - stokes equations .
l. demkowicz and j. t. oden .
an adaptive characteristic petrov - galerkin finite element method for convection - dominated linear and nonlinear parabolic problems in two space variables .
, 55(1 - 2):6387 , apr .
1986 .
j. douglas , jr . and t. f. russell . numerical methods for convection - dominated diffusion problems based on combining the method of characteristics with finite element or finite difference procedures .
, 19(5):871885 , 1982 .
i. sorbents , k. u. totsche , p. knabner , and i. kgel - knabner .
the modeling of reactive solute transport with sorption to mobile and immobile sorbents , 1 . experimental evidence and model development .
, 32:16111622 , 1996 . | in this paper , we consider the adaptive eulerian lagrangian method ( elm ) for linear convection - diffusion problems . unlike the classical a posteriori error estimations
, we estimate the temporal error along the characteristics and derive a new a posteriori error bound for elm semi - discretization . with the help of this proposed error bound ,
we are able to show the optimal convergence rate of elm for solutions with minimal regularity . furthermore , by combining this error bound with a standard residual - type estimator for the spatial error , we obtain a posteriori error estimators for a fully discrete scheme .
we present numerical tests to demonstrate the efficiency and robustness of our adaptive algorithm .
[ multiblock footnote omitted ] [ multiblock footnote omitted ] |
the most metal - poor stars , believed to have formed at redshifts @xmath7 5 ( see e.g. * ? ? ? * ) , and representing well - defined points in space and time , hold clues on the conditions at the earliest epochs that are provided by no other astronomical objects . that is to say , the study of their metallicity distribution function ( mdf ) , together with their relative chemical abundance patterns , have the potential to shed light on the nature of the first generation of objects to form in the universe , and the manner in which the stellar - made elements ( those heavier than li ) first formed .
consider first the mdf .
four decades after the classic analysis of the archtypical metal - poor stars hd 19445 and hd 140283 by @xcite , it could be claimed ( e.g. * ? ? ?
* ) that the mdf for halo material with = log(n(fe)/n(h))@xmath8 log(n(fe)/n(h)@xmath9 , @xmath10 = log(n(fe)/n(h))@xmath8 + 12.00 ] [ fe / h ] @xmath7 4.0 was in reasonable agreement with the predictions of the simple closed box model of galactic chemical enrichment ( the simple model ) of @xcite and @xcite .
following efforts since that time , however , it has become clear that this is not the case below [ fe / h ] @xmath4 4.0 : while two objects are currently known at [ fe / h ] @xmath5 5.3 ( he 01075840 , @xcite , and he 1327@xmath12326 , @xcite ) , the mdf presented by @xcite and the simple model lead one to expect some 40 such objects below [ fe / h ] = 4.0 . figure [ fig : lowfe_mdf ] , which shows the mdf for the unbiased metal - poor objects that have been the subject of high - resolution , high signal - to - noise ( @xmath0 ) model - atmosphere ( one - dimensional ( 1d ) , local thermodynamic equilibrium ( lte ) ) chemical abundance analyses @xmath4 2.0 .
the low spectral - resolution prism surveys , which supply the vast majority of stars known with [ fe / h ] @xmath4 3.0 , are unable to detect the presence of the ca ii k line for all but the very coolest stars . as a result ,
the possibility of introducing bias into the selection function for the lowest metallicity stars is even more remote .
when high - resolution spectra are available one is able to guard against contamination by stellar ch and interstellar ca ii lines that might have militated against recognition of the most metal - poor objects in the discovery phase .
see @xcite . ] , demonstrates a second tantalizing fact .
in contrast to the expectation of a continuously decreasing mdf , the two objects at [ fe / h ] @xmath5 5.3 lie some 1.5 dex below the next most metal - poor objects at [ fe / h ] = 4.0 . despite low numbers ,
this has led some ( e.g. * ? ? ?
* ) to speak of the possibility of a metallicity gap " at lowest abundance , which could have profound implications for our understanding of the nature of the first generation of objects that chemically enriched the galaxy , and presumably the universe .
consider next the relative abundance characteristics of the two stars with [ fe / h ] @xmath5 5.3 .
their most distinctive features are the enormous overabundances , relative to iron , of carbon ( [ c / fe ] @xmath5 4 ) , nitrogen ( [ n / fe ] @xmath5 24 ) , and oxygen ( [ o / fe ] @xmath5 23 ) ( see also @xcite and @xcite ) .
this , taken together with the well - established fact that the incidence of carbon richness increases as one proceeds to lower abundance in the range 4.0 @xmath4 [ fe / h ] @xmath4 2.0 @xcite , suggests that the trend of increasing carbon - richness continues to ubiquitous c enhancement at [ fe / h ] @xmath5 5.0 .
figure [ fig : lowfe_mdf ] shows this trend , where the shaded regions represent objects with [ c / fe ] @xmath7 1.0
. the simplest explanation of this phenomenon is that the earliest objects in the universe produced large amounts of cno relative to fe , far in excess of the ratios produced at later times .
candidate sites include supernovae with mixing and fallback " ( see @xcite , and references therein ) and rotating massive stars @xcite .
@xcite argue that relative overabundances of carbon ( and/or oxygen ) are necessary below [ fe / h ] = 4.0 to provide the cooling of primordial material via and at the earliest times to form the low mass stars we observe today .
we shall return to this point in [ cno ] two questions then beg to be addressed .
first , is the 5.3 @xmath4 [ fe / h ] @xmath4 4.0 gap " real , and second , what is the incidence of cno richness below [ fe / h ] @xmath4 4.0 ?
given the rarity of objects below this limit , one might expect definitive answers to be some time in coming .
that said , the purpose of the present paper is to begin the task .
we present here the discovery and analysis of he 0557@xmath14840 , a third object having [ fe / h ] @xmath4 4.0 . in [ observations ]
we report its discovery , and the high - resolution , high @xmath0 data obtained to address these questions .
our model - atmosphere analysis of this material to produce accurate chemical abundances is then described in [ parameters][analysis ] .
finally , in [ discussion ] we discuss the significance of this object . while much future work remains to be done , we find that the existence of he 0557@xmath14840 with [ fe / h ] = 4.75 weakens the case for the 5.3 @xmath4 [ fe / h ] @xmath4 4.0 gap " canvassed above . on the other hand , with [ c / fe ] = + 1.6 ( and [ n / fe ] @xmath4 + 1.5 and
[ o / fe ] @xmath4 + 3.1 ) , the data for he 0557@xmath14840 are consistent with the suggestion that below [ fe / h ] @xmath4 4.0 all stars possess strong enhancement of at least some elements of the cno group .
we summarize our results in [ summary ] .
he 0557@xmath14840 was observed in 2005 december during an observing session on candidate metal - poor stars from the hamburg / eso objective - prism survey ( he s ; * ? ? ?
* ) with the australian national university s 2.3 m telescope / double beam spectrograph combination on siding spring mountain . the spectra were observed with resolving power r @xmath5 2000 and cover the wavelength range 36005400 .
comparison of the spectrum of he 0557@xmath14840 with standard " metal - poor stars also observed during the program immediately revealed it as an object of considerable interest .
figure [ fig : he0557_lowres ] shows the spectra of he 0557@xmath14840 and of cd@xmath11 , the well - known ultra - metal - poor red giant ( @xmath6 } = -4.0 $ ] ; @xcite , @xcite , @xcite ) .
comparison of these spectra shows that the two stars have comparable hydrogen line strength , and therefore similar effective temperature , while the ca ii line is weaker in he 0557@xmath14840 , suggestive of a lower abundance of calcium .
perhaps equally interesting is that the g band of ch at 4300 is considerably stronger in he 0557@xmath14840 . given that ch is not detected in cd@xmath11 ( [ c / fe ] @xmath4 0.3 , @xcite ) it was immediately obvious that he 0557@xmath14840 is carbon rich relative to cd@xmath11 .
this object , with @xmath12h @xmath13 m @xmath14s and @xmath15 , and @xmath16 = 15.45 and @xmath17 = 0.71 ( see [ photometry ] ) , was clearly a target of interest for high resolution follow - up spectroscopy .
he 0557@xmath14840 was then observed in service mode at the very large telescope ( vlt ) unit telescope 2 ( ut2 ) with the ultraviolet - visual echelle spectrograph ( uves ) during the nights of 2006 february 5 , 25 , and 26 and march 19 .
seven individual exposures with an integration time of 1h each were obtained .
uves was used in dichroic mode with the blue390 and red580 settings .
the useful wavelength ranges , in the rest frame of he 0557@xmath14840 , are 33004520 in the blue - arm spectra , and 48205750 and 58356800 in the lower- and upper- red - arm spectra , respectively .
a @xmath18 slit was used in both arms , yielding a nominal resolving power of @xmath19 .
since read - out noise was not a limitation , @xmath20 pixel binning was employed during the observations , even though this results in a significant oversampling due to the scale of @xmath21 per pixel of the blue - arm ccd and @xmath22 per pixel in the red - arm .
the seven pipeline - reduced spectra were co - added in an iterative procedure in which pixels in the individual spectra affected by cosmic ray hits not fully removed during the data reduction , affected by ccd defects , or other artifacts , were identified .
these pixels were flagged and ignored in the final iteration of the co - addition .
the co - added blue - arm spectra , rebinned by a factor of 2 , have a maximum @xmath0 per pixel of @xmath23 at @xmath24 , decreasing to @xmath25 at the red end . at the blue end , this decreases to @xmath26 at 3700 , and @xmath27 at 3300 . in the rebinned lower red - arm spectra ,
the @xmath0 increases linearly from @xmath28 per pixel at 4820 to @xmath29 at 5750 .
it is approximately constant at a level of @xmath30 per pixel throughout the wavelength range covered by the upper red - arm spectra .
equivalent widths for he 0557@xmath14840 were measured independently by the first two authors , using techniques described by @xcite and @xcite , for unblended lines defined in those investigations , together with that of @xcite .
a comparison of the two data sets is presented in figure [ fig : eqwtesthe0557 ] , where the agreement is quite satisfactory , with an rms scatter about the line of best fit of 2.5 m . in the absence of an understanding of the origin of the small departure from the one - to - one line evident in the figure ,
we have chosen to average the data , and present results for 100 lines suitable for abundance analysis in column ( 5 ) of table [ tab : linelist ] .
the atomic data in the table were mostly taken from the vienna atomic line database ( vald ; @xcite , @xcite ) , complemented where necessary by values from the three works cited above .
for completeness , we identify in table [ tab : rejectedlines ] of appendix a a further 45 lines that we regarded as unsuitable for abundance analysis .
heliocentric radial velocities for he 0557@xmath14840 were measured for the seven individual spectra described above by gaussian fits of six clean and moderately strong absorption lines .
the resulting values are presented in table [ tab : he0557_velocities ] where columns ( 1 ) and ( 2 ) contain the modified julian date and heliocentric velocity , respectively .
the mean internal standard error of measurement of a single observation is 0.2 kms@xmath31 , while the external error is estimated to be 0.7 kms@xmath31 @xcite . given that the dispersion of velocities in the table is 0.21 kms@xmath31 , we conclude that to within the accuracy of our measurements the velocity of he 0557@xmath14840 did not vary over the 43 days of the present observations .
it is interesting to note the large radial velocity , 211.8 kms@xmath31 , of he 0557@xmath14840 , in comparison with values of 44.1 kms@xmath31 for he 0107@xmath15240 @xcite and 63.8 kms@xmath31 for he 1327@xmath12326 @xcite .
the large spread in these three values establishes that stars with [ fe / h ] @xmath4 4.0 have a velocity dispersion commensurate with that of more metal - rich halo material ( [ fe / h ] @xmath7 4.0 ) . in table
[ tab : he0557_photometry ] , we present photometry for he 0557@xmath14840 .
the values for the @xmath32 photometry were taken from @xcite , while the @xmath33 magnitudes are from @xcite . in this table
we also list data for he 01075840 ( [ fe / h ] = 5.3 ) from @xcite , for comparison purposes .
for the reddening of he 0557@xmath14840 we adopt @xmath34 = 0.04 following @xcite , and the differential relationships @xmath35 and @xmath36 following @xcite . for he 0107@xmath15240 we use @xmath34 = 0.013 .
following @xcite , we employ the calibrations of @xcite and @xcite to determine effective temperatures for both he 0557@xmath14840 and he 0107@xmath15240 .
our results are presented in table [ tab : he0557_teff ] , where the errors correspond to uncertainties of 0.02 , 0.02 , and 0.06mag in @xmath37 , @xmath38 , and @xmath39 , respectively , based on the errors in photometry cited by @xcite and @xcite , and reasonable errors in reddening .
weighting the values from the three colors in the table by the inverse square of their errors , we find photometric temperatures @xmath40k and @xmath41k for he 0557@xmath14840 and he 0107@xmath15240 , respectively .
the present results for he 0107@xmath15240 agree well with those of @xcite .
we note that the @xmath42 and @xmath43 colors indicate insignificant effective - temperature differences between he 0557@xmath14840 and he 0107@xmath15240 , while the two calibrations of @xmath44 point towards an effective - temperature difference of around @xmath45k between the two objects .
the source of this difference is not understood . for completeness
, we also note that for he 0557@xmath14840 @xmath46 @xmath1 @xmath47 @xmath5 @xmath48k , while for he 0107@xmath15240 the corresponding quantity is @xmath49k .
to constrain the effective temperature further , we compare the observed profiles of h@xmath50 and h@xmath51 with synthetic profiles based on the recipes described in @xcite .
while this implementation does not take into account the latest advances in the modelling of the self - broadening of balmer lines @xcite , the use of that theory would lower the effective temperature by no more than 50k , as is evident from the analysis of he 0107@xmath15240 @xcite .
even smaller differences are obtained when the effective temperatures are determined relative to the @xmath2 value derived for the solar balmer lines in a differential analysis .
h@xmath50 ( see figure [ fig : halpha ] ) and h@xmath51 unanimously point towards an effective temperature of 4900 @xmath52 100k , lending support to the photometric calibrations giving a large ( @xmath5 @xmath45k ) effective - temperature difference between he 0557@xmath14840 and he 0107@xmath15240 ( see column ( 6 ) of table [ tab : he0557_teff ] ) . while the line strength of h@xmath50 directly correlates with gravity , the opposite is true for h@xmath51 and the higher - order balmer lines .
in addition to determining @xmath53 , the combined analysis of h@xmath50 and h@xmath51 allows one to set constraints on @xmath3 as well ( see [ logg ] )
. given this difference in photometric and spectroscopic effective temperatures , in what follows we shall present abundances for he 0557@xmath14840 determined for both @xmath54k and @xmath55k . using 1d model atmospheres described in
[ atmospheres ] below and assuming lte , we determined the microturbulence velocity @xmath56 from 60 lines by requiring that there be no trend of fe abundance with line strength .
all lines below 3700 were ignored , because the @xmath0 drops rapidly towards shorter wavelengths , resulting in a higher uncertainty of the equivalent width measurements and an increased line - to - line scatter of the abundance of up to 0.8dex .
the average abundances are @xmath57dex and @xmath58dex for @xmath59k and @xmath55k , respectively .
, for example , reveals that for @xmath60k @xmath61dex / ev .
we have encountered this effect in our analyses of other metal - poor giants , and acknowledge it as a shortcoming of the present 1d analysis . ] in both cases , @xmath62 = 1.8 kms@xmath31 .
we also detect two very weak lines of in our red - arm vlt / uves spectra .
the averaged equivalent widths are 4m for the line at @xmath63 and 8m for the line at @xmath64 . for the latter line
, however , spectrum synthesis has to be used to derive an abundance , because the line is blended with a weak line at @xmath65 , having excitation potential @xmath66ev and @xmath67 according to vald . taking the blend into account and using atomic data from vald ( see table [ tab : linelist ] )
, the abundances inferred from the two lines are discrepant by @xmath68dex for @xmath54k .
this difference can be partly removed by using `` astrophysical '' @xmath69 values for the two lines .
that is to say , the @xmath69 values are determined by fitting the solar spectrum , resulting in values of @xmath70 and @xmath71 for @xmath63 and @xmath64 , respectively , where following @xcite we adopt @xmath10 = 7.45 .
the abundance difference is then reduced to @xmath72dex .
there is also considerable uncertainty in the measurement of the equivalent widths of these lines , because they are weak , and therefore small changes in the continuum placement have a large effect on the measured line strength .
furthermore , usage of different fitting methods ( e.g. , using a fixed line width when fitting a gaussian profile to the line as opposed to leaving the line width as a free parameter during the fit ) result in measurements deviating significantly from each other .
the uncertainty is illustrated by the fact that the measurements of the @xmath63 line by a.j.k .
( 6.6m ) and n.c .
( 4.4m ) differ by 2.2m , translating to an abundance difference of @xmath73dex . in summary
, the abundance difference of @xmath74dex can be well - explained by the combined uncertainties of the line strength measurements and @xmath69 values for the two lines .
the iron abundance determined from lines is less sensitive to changes of @xmath2 than that from lines , and therefore the former is less affected by possible systematic errors in the adopted @xmath2 scale .
further , lines are less affected by non - lte effects than those of . for these reasons ,
we adopt the abundance determined from the two lines as the fe abundance of he 0557@xmath14840 . establishing
the ionization equilibrium of an element is a standard spectroscopic method of deriving the stellar surface gravity . to constrain the surface gravity of he 0557@xmath14840
, we perform a non - lte analysis using the iron model atom and line data calibrated on local halo stars with good hipparcos parallaxes @xcite . with a mean overionization of @xmath75dex in fe i
, the ionization equilibrium is balanced at @xmath76k ( from h@xmath50 and h@xmath51 ) and @xmath77 = 2.2 .
assuming an effective temperature of @xmath55k , the iron ionization equilibrium would require a much higher gravity , of the order of @xmath3 = 3 .
such a stellar - parameter combination is in conflict with both the gravity derived from an appropriate isochrone following the procedure of @xcite for he 0107@xmath15240 , one obtains log@xmath78 = 2.2 and the constraints of balmer - profile fitting ( see figure [ fig : halpha ] ) . as we shall see in [ calcium
] the ca i / ca ii ionization equilibrium at @xmath2 = 4900k is also consistent with @xmath3 = 2.2 .
for various reasons ( use of a slightly different model atmosphere and @xmath79 values , different line selection etc . )
the lte ionization equilibrium was established by n.c . at @xmath3 = @xmath80
; that is , at a higher gravity than indicated by the non - lte analysis , but without any consideration of overionization .
the combination of balmer - profile temperatures and fe i / fe ii non - lte surface gravities was recently shown to yield stellar parameters for metal - poor giant stars in excellent agreement with stellar evolution @xcite .
nonetheless , given the uncertainties arising from the unknown absolute temperature scale of metal - poor stars , we have to accept a potential systematic error in absolute @xmath3 values of up to 0.3dex .
table [ tab : stellarparameters ] summarizes our atmospheric parameters for he 0557@xmath14840 , based on spectroscopic and photometric data .
while we consider the purely spectroscopic stellar parameters to be more trustworthy , abundance ratios for both solutions are presented for the reader s convenience in what follows . the iron abundance for he 0557@xmath14840 is [ fe / h ] = 4.75 or 4.71 , depending on which of the spectroscopic or photometric solutions , respectively , one prefers .
we have computed 1d model atmospheres tailored for the abundances of he 0557@xmath14840 ; i.e. , taking into account in particular the over - abundance of c , as determined during a next - to - final iteration of our abundance analysis . an enhancement of the @xmath50-elements , including oxygen , of @xmath81dex , was assumed .
the models were computed with the marcs code , using opacity sampling in about 100,000 wavelength points .
atomic and diatomic molecular opacities were included in addition to the continuous ones in the model atmosphere computations .
the models are hydrostatic , spherically symmetric and computed under the assumption of lte ; convective energy transport is included using a mixing length theory formulation .
further details of the model atmospheres can be found in @xcite .
following @xcite , we computed lte abundances for the transitions in table [ tab : linelist ] , using the atomic data and equivalent widths ( and their limits ) presented there , the atmospheric parameters of table [ tab : stellarparameters ] , and the code ` eqwi ` , version 7.04 .
the results are presented in columns ( 6 ) and ( 7 ) of table [ tab : linelist ] , for @xmath54k and @xmath55k , respectively .
for a few lines ( in particular those for which only an upper line strength limit appears in table [ tab : linelist ] ) spectrum synthesis was performed with the code ` bsyn ` , version 7.04 .
figure [ fig : srii4077synth ] presents an example of spectrum synthesis in the region of the sr ii 4077.71 line , which was used to confirm the upper limit to the strontium abundance derived by the equivalent width analysis .
spectrum synthesis was also used in the analysis of lines of , , , and ( see [ micro ] ) . , with numerous weak lines in the range
@xmath5 34003600 ( and hence difficult to measure in he 0557@xmath14840 , given the @xmath0 of the spectra ) , was analysed by using spectrum co - addition for several lines as described by @xcite in their analysis of oh lines in he 1300 + 0157 .
( in this method several lines are chosen , the wavelength scale is set to zero at line centre for each line , and the spectra co - added to yield a composite spectrum . ) our adopted application of the technique is shown for the co - addition of the strongest two lines in figure [ fig : coaddco ] , which presents a comparison of observed and synthetic spectra of he 0557@xmath14840 and cd@xmath11 .
( for he 0557@xmath14840 the synthetic spectra pertain to @xmath2 = 4900k , while for cd@xmath11 we adopt @xmath54k , log@xmath78 = 2.0 following @xcite , @xmath62 = 2.4 kms@xmath31 , @xmath82 = 3.39 from @xcite , and @xmath83 = 1.44 , and hence [ co / fe ] = 0.58 obtained for the two co - added lines ( christlieb , priv .
comm . ) ) . for he
0557@xmath14840 and cd@xmath11 figure [ fig : coaddco ] yields [ co / fe ] = 0.04 and 0.7 , respectively , with the latter in good accord with the result of christlieb et al . cited above .
our final results for the atomic lines of 18 elements ( ten detections , eight upper limits ) are presented in table [ tab:1dlteabundances ] .
assuming lte , we have used spectrum synthesis techniques to seek to analyze features of ch , nh , and cn and constrain the abundances of carbon , nitrogen , and oxygen .
the c abundance of he 0557@xmath14840 was determined from weak to moderately strong ch a - x lines in the wavelength region 42524256 ( see figure [ fig : ch4250fit ] ) , the head of the g band at @xmath84 , the feature at @xmath85 ( see figure [ fig : ch4323fit ] ) , and weak lines of the b - x system in the region around k. we used a line list compiled by b. plez and a. jorisson ( 2006 , priv .
the @xmath69 values and line positions were taken from lifbase @xcite , and the excitation energies from @xcite .
the abundances all agree to within 0.05dex with each other , and we adopt the average of the results , which is listed for both combinations of the stellar parameters in table [ tab:1dlteabundances ] .
we determined an upper limit for the nitrogen abundance of he 0557@xmath14840 by analyzing the @xmath86 band head of the b - x system of cn at 3883 ( see figure [ fig : cn3883synth ] ) . in these calculations
we use a line list of b. plez ( 2001 , priv .
the details of its computation are described in @xcite .
we adopt a dissociation energy for the cn molecule of @xmath87ev , and the c abundance obtained from the ch lines .
stronger upper limits can be inferred from the absence of the @xmath86 band head of the a - x system of nh at @xmath88 ( see figure [ fig : nhsynth ] ) , yielding @xmath89 and @xmath90 for @xmath54k and @xmath91k , respectively . for our spectrum synthesis calculations we used the line list of @xcite with the @xmath69 values reduced by @xmath74 , and a dissociation energy for nh of @xmath92ev . for a discussion of the reasons for the correction of the @xmath69 values and the choice of @xmath93 see @xcite . in table [ tab:1dlteabundances ] , we have listed upper limits for the oxygen abundance , derived from the upper limit to the line strength of the line at 6300.30 . due to our vlt / uves spectrum beginning only at 3300 it was not possible to determine the oxygen abundance from the uv oh a - x lines . as well as iron , calcium is also observed in two ionization stages .
this offers the possibility of checking the spectroscopic stellar - parameter determination , primarily the gravity , independently .
as can be seen from the three calcium - line entries of table [ tab : linelist ] , the two ionization stages indicate significantly different lte abundances . for the @xmath2 = 4900k case , @xmath94 ( ca i 4226 ) = @xmath95 and @xmath94 ( ca ii 3933 ) = @xmath96 .
the abundance derived from the subordinate line ca ii 3706 falls in between
@xmath94 ( ca ii 3706 ) = @xmath97 with some added uncertainty arising from blends with ch .
the situation is similar for the @xmath2 = 5100k solution , albeit at higher calcium abundances . according to computations by @xcite , both ca
i 4226 and ca ii 3706 are affected by non - lte .
such corrections raise the abundances derived from these lines by @xmath5 + 0.30 and + 0.17dex , respectively .
the non - lte corrections are similar for both solutions . while this brings the two observed ca ii lines into excellent agreement for the @xmath2 = 4900k solution , the predicted non - lte effect for ca i 4226 is too small to establish the ionization equilibrium ( by 0.2dex ) .
thus , while the two lines of ca ii can be reconciled at 4900k , the ionization equilibrium seems to ask for a significantly lower surface gravity than log@xmath78 = 2.2 . for the @xmath2
= 5100k solution , the non - lte abundance of ca i 4226 is 0.07dex larger than that derived from ca ii 3706 .
this indicates that the gravity would have to be raised by about 0.3dex to establish the ionization equilibrium . at these stellar parameters ,
ca ii 3933 would yield an abundance 0.18dex above the other two lines .
the corresponding trend in abundance versus excitation energy in ca ii is reminiscent of similar trends seen in fe i for metal - poor stars including he 0557@xmath14840 and could indicate deficits in the t@xmath98 relation of hydrostatic model atmospheres .
table [ tab : calcium ] indicates how the situation would change , for @xmath2 = 4900k , if the analysis were done differentially with respect to the sun employing the methodology and @xmath79 values favored by @xcite .
column ( 2 ) presents the non - lte corrections discussed above , while columns ( 3 ) and ( 4 ) give the changes due to different @xmath79 values , and fitting to the sun , respectively .
( since both lines of ca ii are very strong and heavily blended in the solar spectrum , no attempt was made to derive astrophysical @xmath79 values for them . in these cases ,
only the corrections given in columns ( 2 ) and ( 3 ) were employed . )
the final calcium abundances are presented in column ( 5 ) . as can be seen ,
the ionization equilibrium of ca i 4226 and the @xmath77-sensitive line ca ii 3706 can be established to within 0.03dex at @xmath2 = 4900k .
the three lines then indicate a calcium abundance @xmath94 ( ca ) = @xmath99 @xmath52 0.03 ( 1@xmath100 ) . for the @xmath2 = 5100k solution , the abundance from ca
i 4226 is 0.25dex larger than that from ca
ii 3706 . this would require a substantially ( @xmath5 1dex ) larger @xmath3 ( cf .
[ logg ] ) .
accordingly , the abundances from the three lines scatter significantly more @xmath101 ( ca ) = @xmath102 @xmath52 0.14 , if @xmath103=2.2 is assumed .
we conclude that , within the uncertainties of the present analysis , it is possible to establish the fe i / fe ii and ca i / ca ii ionization equilibria at @xmath2 = 4900k and @xmath3 = 2.2 , while this is not the case for @xmath2 = 5100k and @xmath3 = 2.2 .
adding balmer - profile analyses to the picture , we therefore favour the @xmath2 = 4900k solution .
we note that the chosen stellar - parameter determination thus employs techniques very similar to those used in the analysis of he 0107@xmath15240 @xcite , putting the chemical abundances of both stars on the same abundance scale and making possible a differential analysis .
abundance errors , within the 1d , lte formalism adopted here , have been discussed in considerable detail by @xcite for he 0107@xmath15240 , to which we refer the reader . given the similarity of temperatures and gravities of that object and he 0557@xmath14840 , we shall not repeat the discussion here , except to note that random abundance errors introduced by uncertainties of the measurements of equivalent widths in a spectrum of the quality that we use here are typically of the order of @xmath104dex , and those caused by uncertainties of atomic data are typically of the order of @xmath105@xmath106dex .
systematic errors of the stellar parameters , most notably @xmath2 , can easily result in systematic difference in the _ absolute _ abundances of 0.2dex , while these errors mostly cancel out for abundance _
ratios _ of atomic species of the same ionization state having similar ionization potentials ( e.g. , [ / ] , or [ / ] ) non - lte corrections are metallicity dependent and can become quite large in extremely metal - poor stars ( especially giants ) , in particular for resonance lines of photoionization - dominated minority species .
this is , for example , the case for ca i 4226 for which the non - lte correction is @xmath5 + 0.5dex for he 1327@xmath12326 .
smaller corrections are expected for lines arising from transitions within dominant ionization stages . however , for subordinate lines corrections can reach @xmath5 0.2dex ( cf .
ca ii 3706 as discussed in
[ calcium ] ) .
finally , in extremely metal - poor giants , the work of @xcite shows that the use of 3d , lte modeling ( as opposed to our 1d , lte procedure ) will lead to significantly lower abundances than reported here . in column
( 6 ) of table [ tab:1dlteabundances ] we present their results for the corrections @xmath107 = @xmath108
@xmath109 for he 0107@xmath15240 ( which has very similar @xmath2 and @xmath3 to those of he 0557@xmath14840 ) .
roughly speaking , our abundances derived from diatomic molecules are too high by @xmath5 1.0 dex , while for atomic species our values are too large by 0.1 to 0.5 , with most values in the range 0.1 0.3 .
that said , one should note the caveat of @xcite in the context of the iron peak elements : we caution ... that the neglected non - lte effects might actually be substantial ... "
while table [ tab:1dlteabundances ] presents abundances for both @xmath54k and @xmath91k , for reasons stated above , our preference is for the former .
accordingly , we shall use results for @xmath54k in what follows . with [ fe / h ] = 4.75
, he 0557@xmath14840 falls midway between the most metal - poor objects at [ fe / h ] @xmath5 5.3 ( he 0107@xmath15240 and he 1327@xmath12326 ) , on the one hand , and the several next metal - poor stars at [ fe / h ] 4.0 , on the other .
_ our first conclusion is that the discovery of he 0557@xmath14840 weakens the case for a gap " in the metallicity distribution function ( mdf ) between [ fe / h ] = 4.0 and 5.3_. we also note that although the search for metal - poor stars in the he s is not yet complete , the survey has already increased the total number of known metal - poor stars by more than a factor three with respect to all previous surveys combined for such objects , and all three stars currently known to have @xmath6 } < -4.0 $ ] were discovered in the he s .
this suggests that the previously suspected absence of stars at @xmath6 } < -4.0 $ ] , as well as the presence of the putative gap " in the mdf were statistical artifacts caused by too small samples of metal - poor stars .
if this explanation is correct , it is expected that the follow - up observations of the remaining he s metal - poor candidates , and even deeper surveys , such as segue ( see e.g. ` http://www.sdss.org/segue ` and the description in * ? ? ?
* ) , will further populate the `` gap '' , as well as the metallicity region below @xmath6 } = -5.0 $ ] .
the shape of the low - metallicity tail of the mdf of the galactic halo contains information on the earliest phases of its chemical enrichment .
@xcite , for example , compares a preliminary mdf derived from the results of spectroscopic follow - up observations of candidate metal - poor stars from the hk survey ( see e.g. @xcite for details of the survey ) with three different models of galactic chemical evolution .
he finds that the low - metallicity tail is reproduced well by a `` simple outflow model '' with a relaxed instantaneous recycling approximation ( ira ) and the assumption of an early phase of infall of primordial gas lasting about 0.2gyr .
the latter assumption reduces the number of stars at @xmath110 } < -3.0 $ ] relative to `` simple outflow '' models without early infall and with or without ira .
such a reduction is adopted to match the observed mdf .
the model predicts , however , an mdf increasing continuously from the lowest [ fe / h ] to the peak of the mdf at around @xmath110}=-1.6 $ ] . from figure 1 of @xcite
, one sees that the number of stars in the range @xmath111 } < -4.0 $ ] should be about five times as high as the number having @xmath110 } < -5.0 $ ] .
since there are now two stars known with @xmath110 } < -5.0 $ ] , about ten stars should have been found in the previously suspected `` gap '' of the mdf while in reality only one has been identified ( he 0557@xmath14840 ) .
similar problems arise for the galaxy formation and chemical evolution model of @xcite . a critical metallicity @xmath112 for low - mass ( i.e. , @xmath113 ) star formation as low as @xmath114 would have to be adopted in order to reproduce the number of stars at @xmath110 } < -5.0 $ ] ( i.e. , two ) correctly , but this would result in a large number of objects in the regime @xmath111 } < -4.0 $ ] , in contradiction with the observations .
@xcite considered a stochastic chemical enrichment model for the galactic halo in which negative feedback effects , from the first generation of ( massive ) stars that formed from primordial gas clouds , resulted in a suppression of all star formation .
later , at redshifts of about @xmath115 , the second generation of stars formed from gas pre - enriched to levels of @xmath110 } < -5.0 $ ] by the previous generation of stars that exploded as type - ii supernovae ( sn ii ) . ( the resulting mass function had a bias towards high mass ( short - lived ) stars except where carbon was sufficiently overabundant with respect to iron to allow low mass ( long - lived ) objects to form . )
the mdf resulting from this model has one ( or more ) `` pre - enrichment '' peaks in the range @xmath116 } < -5.0 $ ] , with the number of stars then first decreasing with increasing [ fe / h ] , until it rises again at @xmath110 } > -4.0 $ ] .
the mdf of karlsson matches the observations reasonably well . in his most favourable case , with suppression of star formation in carbon - deficient gas
, some 34 objects are predicted in the range 5.1 @xmath4 [ fe / h ] @xmath4 4.2 .
the observational constraints on the exact form of the mdf at the lowest metallicities are , however , still weak , since currently only three stars having @xmath110 } < -4.0 $ ] are known .
ongoing and future surveys for metal - poor stars , expected to lead to the identification of significant numbers of new stars at the lowest metallicities as noted above , should clarify the situation .
a particularly interesting and exciting prediction of karlsson s model is that there is an additional , smaller peak in the mdf at around @xmath110 } = -7.0 $ ] . from his figure 2
, one can estimate that among ten stars at @xmath110 } < -5.0 $ ] there should be one at @xmath110 } \sim -7.0 $ ] = 5.7 .
we refer the reader also to the discussion by @xcite who considered possible maximal accretion onto he 0107@xmath15240 , and concluded that most heavy elements of he 0107@xmath15240 _ might _ be accreted from the interstellar medium , although this is most probably not the case for c , n , and na " . what remains to be addressed
is whether minimal accretion conditions existed that would permit one to recognize objects that initially had [ fe / h ] = 7.0 . ]
figure [ fig : relative_abundances ] presents [ x / fe ] as a function of [ fe / h ] for 12 representative elements ( x ) in he 0557@xmath14840 and the other @xmath5 50 metal - poor stars with [ fe / h ] @xmath4 3.0 for which high - resolution , high @xmath0 abundance analyses exist .
the comparison data have been taken with first preference from @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , and @xcite , supplemented by earlier material from @xcite and @xcite . in the figure , he 0557@xmath14840 is represented by a filled star , he 1300 + 0157 ( frebel et al . ) by an open one , results of cayrel et al . and franois et al . by filled circles , and others by open ones .
it is important to note that all of the data in the figure have been determined using 1d model atmospheres and the assumption of lte .
this should be borne in mind when comparison is made with predictions of stellar evolution and galactic chemical enrichment models .
we note in passing that in figure [ fig : relative_abundances ] there are six pairs of related elements representing the cno group , the light odd - numbered elements , the @xmath50-elements , the fe - peak below ( cr and mn ) and above ( co and ni ) iron , and the heavy neutron - capture elements .
as has been discussed at some length in a number of the works cited here , and is clear in the figure , the case can be made that for [ fe / h ] @xmath4 3.0 , and elements with atomic number less than those of the heavy neutron - capture elements , there is a majority fraction that exhibits well - defined trends with small dispersion of [ x / fe ] , together with a minority , c - rich , one that becomes more prominent as one proceeds to lowest abundance and exhibits large enhancements , to varying degrees , of the other lighter elements ( n , o , na , mg , and al ) .
inspection of the cno data for he 0557@xmath14840 leads to our second important conclusion . _ with [ c / fe ] = + 1.65 , [ n / fe ] @xmath4 + 1.47 , and [ o / fe ] @xmath4 + 3.09 , he 0557@xmath14840 belongs to the c - rich group . currently all objects with [ fe / h ] @xmath4 4.0 share this property . _
the question of relative carbon and oxygen overabundance at lowest [ fe / h ] has been discussed by @xcite , who address the difficulty of forming low - mass stars , such as he 0107@xmath15240 and he 1327@xmath12326 , in the early universe .
they argue that cooling by c and o ( via fine - structure lines of and ) provides the means of forming objects with m @xmath4 1 m@xmath9 .
they predict that the formation of such low - mass stars will occur due to cooling by c and o if the
transition discriminant " d@xmath117 satisfies the condition : @xmath118 } } + 0.3\times 10^{\scriptsize \mbox{[o / h]}}\right)\mbox{\ $ > $ --3.5,}\ ] ] as may be seen from their figure 1 , both he 0107@xmath15240 and he 1327@xmath12326 ( with d@xmath119 2.7 to 2.4 , respectively ) easily satisfy the condition .
for he 0557@xmath14840 we have a relative overabundance of carbon but only a weak constraint on that of oxygen .
if we assume that the latter lies in the range zero to @xmath120 5.5 5.0 and 6.0 in he 0107@xmath15240 and he 1327@xmath12326 , respectively .
] , we obtain 4.1 @xmath121 3.5 for @xmath54k .
( we adopt @xmath122 , following @xcite for he 0107@xmath15240 . )
this range lies close to the limit , but is not inconsistent with the hypothesis of @xcite .
that said , more work is clearly needed to further constrain the abundances of o in he 0557@xmath14840 .
a low value for @xmath123 ( o ) could marginally challenge their prediction .
closer inspection of figure [ fig : relative_abundances ] leads to a less exotic , but nevertheless important conclusion .
_ for all elements other than c , with the possible exception of co , he 0557@xmath14840 has [ x / fe ] values that differ little from those found in the majority fraction described above . _ in this regard he 0557@xmath14840 is similar to he 1300 + 0157 ( [ fe / h ] = 3.88 , @xcite ) , represented in the figure by the open star .
given the combination of carbon enhancement and low abundances of the heavy neutron - capture elements , it is evident that he 0557@xmath14840 belongs to the cemp ( carbon enhanced metal poor)-no group ( see * ? ? ?
* ) , which have carbon enhancement and [ ba / fe ] @xmath4 0.0 ( as does he 1300 + 0157 ) . [ ironpeak ]
we draw the reader s attention to the apparently low value of [ co / fe ] for he 0557@xmath14840 with respect to the bulk of the comparison objects in figure [ fig : relative_abundances ] .
while one sees , for cr / fe and mn / fe , that he 0557@xmath14840 displays the relative underabundances , and for ni / fe the solar value , consistent with the trends for extremely metal - poor stars first reported by @xcite and @xcite , co / fe appears to behave differently .
that is to say , for he 0557@xmath14840 one sees [ co / fe ] = 0.0 , somewhat lower than the value [ co / fe ] @xmath5 0.5 that pertains to the bulk of objects with [ fe / h ] @xmath4 3.5 .
a word of caution is , however , perhaps in order : inspection of table [ tab:1dlteabundances ] shows that for @xmath91k one obtains [ co / fe ] = 0.3 for he 0557@xmath14840 .
the reader should bear this in mind in the following discussion .
possible explanations of the relative underabundances of cr and mn , the relative overabundance of co , and the solar behaviour of ni / fe for the bulk of extremely metal - poor stars ( which demonstrate the need for a possibly wide range of conditions during the explosions of the first stars ) include higher than canonical explosion energies ( i.e. , hypernovae " , with @xmath7 10@xmath124 ergs ) and smaller mass cuts than usually adopted to produce sufficient fe .
a third relevant effect is variation in y@xmath125 , the electron mole number = @xmath126z@xmath127x@xmath127/a@xmath127 , where z@xmath127 , x@xmath127 , and a@xmath127 are the atomic number , mass fraction , and mass number of each species , respectively .
neutron excess is then defined as @xmath128 = 1 2y@xmath125 ] .
we refer the reader to @xcite and @xcite for a comprehensive discussion of these points .
suffice it here to say that variation of the explosion energy and of the mass cut can each act to produce the observed coupled relative underabundance behaviour of [ cr / fe ] and [ mn / fe ] , on the one hand , and the relative overabundance one of [ co / fe ] , on the other .
it is difficult , then , to see how the departure of the [ co / fe ] value in he 0557@xmath14840 from the general trend might be explained in this manner . in some contrast ,
varying y@xmath125 on the range 0.49950.5005 in an object with m = 25 m@xmath130 , z = 10@xmath131 , and e = 2@xmath13210@xmath133 ergs leads to strong decoupling of the nexus between these elements ( * ? ? ?
* e.g. , figure 4 ) .
we shall return to [ co / fe ] ( and [ cr / fe ] ) in he 0557@xmath14840 in [ popii ] .
given the similarity between the relative abundances of he 0557@xmath14840 and he 1300 + 0157 ( their iron abundance difference @xmath134[fe / h ] = 0.9 notwithstanding )
we refer the reader to the comprehensive work of @xcite who have considered the various possible origins of the latter , and which has great relevance to the present discussion .
the interpretation of the abundance patterns of the most metal - poor stars turns to some extent on their multiplicity .
for example , while no significant radial velocity variations have been observed for either of the two most metal - poor stars he 0107@xmath15240 and he 1327@xmath12326 ( [ fe / h ] @xmath5 5.3 ) , in the case of the former , @xcite have proposed a population iii , binary model in which he 0107@xmath15240 is a first generation zero heavy - element object that has experienced binary mass transfer from an asymptotic giant branch ( agb ) star , together with accretion from the interstellar medium .
a population ii variation on this theme is that he 0557@xmath14840 could be a second- or later- generation object that has experienced mass transfer from an erstwhile more massive ( possibly agb ) companion , leading to enrichment of carbon but not of the heavy neutron - capture elements .
see @xcite for a discussion of this mechanism . in neither case can such models be excluded as an explanation of he 0557@xmath14840 by the current level of non - detection of radial velocity variations as reported in [ radial_velocities ] ,
none has been detected in the present , albeit sparse , data . future continued radial velocity monitoring is needed to potentially reach a definitive conclusion on its multiplicity .
the interpretation of he 0557@xmath14840 with [ fe / h ] = 4.8 , and [ c / fe ] = + 1.6 as a single , population ii , object requires a scenario in which carbon has been overproduced relative to the heavier elements .
while for he 0557@xmath14840 one seeks to understand the large c enhancement , our current lack of knowledge concerning the abundances of n and o suggests that one might also consider a more general solution ( applicable not only to c but also other light elements ) that involves the preferential expulsion , during the explosion of objects with masses greater than @xmath5 10 m@xmath130 , of the outer layers ( in which the lighter elements but not the heavier ones have been synthesized ) , but in which only a small fraction of the inner regions ( where the heavier elements have also been produced ) is expelled .
some suggestions / models that offer promise in this regard are : @xmath135 rotating , zero - heavy - element ( z = 0 ) , very massive ( 250300 m@xmath130 ) objects @xcite @xmath135 failed ( low energy ) , z = 0 , supernovae ( see e.g. @xcite and @xcite ) @xmath135 z = 0 , 2030 m@xmath130 , faint " supernovae ( explosion energies lower than 10@xmath124 ergs ) with mixing and fallback " ( see @xcite and @xcite ) @xmath135 rotating , z = 0 , 60 m@xmath130 stars @xcite in the context of objects forming from the ejecta of such putative objects , one has also to consider its possible admixture with the existing interstellar medium , which may have already been enriched by the ejecta of less exotic objects , among which one might include non - rotating , energetic core - collapse and pair - instability supernovae in the mass range 10 @xmath4 m / m@xmath130 @xmath4300 ( see e.g. @xcite ; @xcite ; @xcite ) . given the approximate nature of the above one dimensional models , it remains to be seen which , if any , of them might be relevant to the enrichment of the material from which he 0557@xmath14840 and the other c - rich most metal - poor stars formed .
that said , the mixing and fallback " models of @xcite reproduce most of the abundance patterns seen in the objects at [ fe / h ] @xmath5 5.3 .
the growing number of objects with [ fe / h ] @xmath4 3.5 that exhibit relative overabundances of the light elements , together with the diversity of their abundance patterns , are now establishing a challenging framework against which to test and constrain such putative theoretical models . in the context of he 0557@xmath14840 , stronger limits on oxygen and nitrogen are now required to more rigorously constrain our understanding of its formation and evolution .
that said , we noted above that @xcite have offered a comprehensive critique of the various models relevant to the manner in which he 1300 + 0157 may have formed , which is useful in informing the present discussion of he 0557@xmath14840 .
their basic conclusion was that the cemp - no abundance pattern of he 1300 + 0157 could perhaps result in either of two ways .
their first scenario envisages enrichment from a single faint " supernova , with mixing and fallback " ( explosion energies lower than 10@xmath124 ergs , see e.g. @xcite and @xcite ) , with an unusual mass cut allowing for excesses in c and o " .
it will be interesting to see whether this versatile type of model can explain the fe - peak data for he 0557@xmath14840 in figure [ fig : relative_abundances ] . certainly the mixing and fallback model of @xcite for the giant he 0107@xmath15240 ( with similar @xmath53 and log@xmath78 to those of he 0557@xmath14840 ) , with [ co / fe ] @xmath5 1.6 , more than overcomes the problem of low [ co / fe ] for he 0557@xmath14840 discussed in [ ironpeak ] .
on the other hand , in that model one finds the prediction of [ cr / fe ] @xmath5 0.3 , considerably higher than [ cr / fe]@xmath136 = 1.1 reported here for he 0557@xmath14840 . ] .
their second suggestion postulates two enrichment sources , with the admixture of ejecta from a massive slowly rotating population iii object ( m @xmath7 100 m@xmath130 , z = 0 ; see e.g. @xcite ) which produces copious amounts of c and o , but little n and heavier elements before exploding as a pair - instability sn , together with material from a hypernova " ( explosion energy e @xmath5 10@xmath13210@xmath124 ergs , see e.g. @xcite , and @xcite ) which would contribute heavier elements having atomic number less than 30 .
there are , however , problems with this model as an explanation of he 0557@xmath14840 . while it is consistent with the observation of [ co / fe ] = 0.76 in he 1300 + 0157 , it faces difficulty as an explanation of [ co / fe ] = 0.0 ( [ co / fe]@xmath136 = 0.3 ) in he 0557@xmath14840 , since , as discussed above in [ ironpeak ] , the hypernovae of @xcite and @xcite produce higher co abundances , for which they report [ co / fe ] @xmath5 0.4 .
we have reported the discovery and analysis of the ultra metal - poor red giant + he 0557@xmath14840 , which has atmospheric parameters @xmath2 = 4900k , @xmath3 = 2.2 , and [ fe / h ] = 4.75 .
this object is also carbon rich , with [ c / fe ] = + 1.6 , a property shared by the three objects currently known to have iron abundance less than [ fe / h ] @xmath5 4.0 .
we have also obtained relative abundances , or abundance limits , for a further 18 elements , and find that , with the possible exception of [ co / fe ] , he 0557@xmath14840 has an abundance pattern similar to that of the majority of objects with [ fe / h ] @xmath5 4.0 .
he 0557@xmath14840 falls in the `` gap '' between the two most metal - poor stars at [ fe / h ] @xmath5 5.5 and the next most metal - poor stars at [ fe / h ] = 4.0 .
its discovery weakens the case for the existence of such a gap " .
( that said , while the statistics are poor , the number of objects with 5.0 @xmath4 [ fe / h ] @xmath4 4.0 is lower by a factor @xmath534 than predicted at the earliest times by existing models of galactic chemical enrichment . ) the abundance pattern of he 0557@xmath14840 has been discussed in the context of single and two - component enrichment models .
existing `` mixing and fallback '' models provide a poor explanation of the observed relative abundances , while a two - component model provides a better , but still inadequate one .
the overabundance of carbon in he 0557@xmath14840 is ( marginally ) consistent with the hypothesis of @xcite that carbon ( and oxygen ) richness may be a necessary requirement for formation of low mass stars at the earliest times .
further data are required to determine the oxygen abundance and improve that of co , and hence more strongly constrain the origin of he 0557@xmath14840 .
we would like to thank the director s discretionary time committee of eso for allocating time to this project .
we express our gratitude to the eso staff on paranal and in garching for carrying out the observations , and reducing the data , respectively .
we are grateful to thomas szeifert for making it possible for the observations of he 0107@xmath15240 to be initiated on the night of 5 february , 2006 , and to b. plez and a. jorissen for providing us with @xmath137ch and @xmath138ch line lists .
we also thank brian schmidt for beneficial discussion on the kinematics of he 0557@xmath14840 .
j.e.n and m.s.b acknowledge support from the australian research council under grants dp0342613 and dp0663562 .
n.c . is a research fellow of the royal swedish academy of sciences supported by a grant from the knut and alice wallenberg foundation .
he also acknowledges financial support from deutsche forschungsgemeinschaft through grants ch 214/3 and re 353/44 .
a.j.k . and k.e .
acknowledge support from swedish research council grants 50349801 and 60307001 , respectively . t.c.b .
acknowledges partial funding for this work from grants ast 04 - 06784 and phy 02 - 16873 : physics frontier center/ joint institute for nuclear astrophysics ( jina ) , both awarded by the national science foundation .
+ + . _
facilities : _ anu : sso2.3m(dbs ) ; vlt : kueyen(uves ) 75 natexlab#1#1 alonso , a. , arribas , s. , & martnez - roger , c. 1998 , a&as , 131 , 209 . 1999 , a&as , 140 , 261 alonso , a. , arribas , s. , & martnez - roger , c. 2001 , a&a , 376 , 1039 aoki , w. , frebel , a. , christlieb , n. , norris , j. e. , beers , t. c. , minezaki , t. , barklem , p. s. , honda , s. , takada - hidai , m. , asplund , m. , ryan , s. g. , tsangarides , s. , eriksson , k. , steinhauer , a. , deliyannis , c. p. , nomoto , k. , fujimoto , m. y. , ando , h. , yoshii , y. , & kajino , t. 2006 , apj , 639 , 897 aoki , w. , norris , j. e. , ryan , s. g. , beers , t. c. , & ando , h. 2002 , apj , 576 , l141 aoki , w. , norris , j. e. , ryan , s. g. , beers , t. c. , christlieb , n. , tsangarides , s. , & ando , h. 2004 , apj , 608 , 971 asplund , m. , grevesse , n. , & sauval , a. j. 2005 , in asp conf .
336 , cosmic abundances as records of stellar evolution and nucleosynthesis in honor of david l. lambert , ed .
t. g. barnes & f. n. bash , san francisco , 25 barklem , p. s. , piskunov , n. , & omara , b. j. 2000 , a&a , 363 , 1091 beers , t. c. & christlieb , n. 2005 , ara&a , 43 , 531 beers , t. c. , christlieb , n. , norris , j. e. , bessell , m. s. , wilhelm , r. , allende prieto , c. , yanny , b. , rockosi , c. , newberg , h. j. , rossi , s. , & lee , y. s. 2005 , in iau symposium , vol .
228 , from lithium to uranium : elemental tracers of early cosmic evolution , ed .
v. hill , p. franois , & f. primas , 175 beers , t. c. , flynn , c. , rossi , s. , sommer - larsen , j. , wilhelm , r. , marsteller , b. , lee , y. s. , de lee , n. , krugler , j. , deliyannis , c. p. , simmons , a. t. , mills , e. , zickgraf , f .- j . ,
holmberg , j. , nehag , a. , eriksson , a. , terndrup , d. m. , salim , s. , andersen , j. , nordstrm , b. , christlieb , n. , frebel , a. , & rhee , j. 2007 , apj suppl .
, 168 , 128 bessell , m. s. , castelli , f. , & plez , b. 1998 , a & , 333 , 231 bessell , m. s. , christlieb , n. , & gustafsson , b. 2004 , apj , 612 , l61 carpenter , j. m. 2001 , aj , 121 , 2851 carretta , e. , gratton , r. , cohen , j. g. , beers , t. c. , & christlieb , n. 2002 , aj , 124 , 481 cayrel , r. , depagne , e. , spite , m. , hill , v. , spite , f. , franois , p. , beers , t. c. , primas , f. , andersen , j. , barbuy , b. , bonifacio , p. , molaro , p. , & nordstrm , b. 2004 , a&a , 416 , 1117 chamberlain , j. w. & aller , l. h. 1951 , apj , 114 , 52 christlieb , n. , bessell , m. s. , beers , t. c. , gustafsson , b. , korn , a. , barklem , p. s. , karlsson , t. , mizuno - wiedner , m. , & rossi , s. 2002 , nature , 419 , 904 christlieb , n. , gustafsson , b. , korn , a. j. , barklem , p. s. , beers , t. c. , bessell , m. s. , karlsson , t. , & mizuno - wiedner , m. 2004 , apj , 603 , 708 2007 , in preparation clarke , c. j. & bromm , v. 2003 , mnras , 343 , 1224 cohen , j. g. , christlieb , n. , mcwilliam , a. , shectman , s. , thompson , i. , wasserburg , g. j. , ivans , i. i. , dehn , m. , karlsson , t. , & mel ' endez , j. 2004 , apj , 612 , 1107 collet , r. , asplund , m. , & trampedach , r. 2006 , apj , 644 , l121 franois , p. , depagne , e. , hill , v. , spite , m. , spite , f. , plez , b. , beers , t. c. , barbuy , b. , cayrel , r. , andersen , j. , bonifacio , p. , molaro , p. , nordstrm , b. , & primas , f. 2003 , a&a , 403 , 1105 frebel , a. , aoki , w. , christlieb , n. , ando , h. , asplund , m. , barklem , p. s. , beers , t. c. , eriksson , k. , fechner , c. , fujimoto , m. y. , honda , s. , kajino , t. , minezaki , t. , nomoto , k. , norris , j. e. , ryan , s. g. , takada - hidai , m. , tsangarides , s. , & yoshii , y. 2005 , nature , 434 , 871 frebel , a. , christlieb , n. , norris , j. e. , aoki , w. , & asplund , m. 2006 , apjl , 638 , l17 frebel , a. , johnson , j. l. , & bromm , v. 2007 , apj , submitted , astro - ph/0701139 frebel , a. , norris , j. e. , aoki , w. , honda , s. , bessell , m. s. , takada - hidai , m. , beers , t. c. , & christlieb , n. 2007 , apj , in press , astro - ph/0612160 fryer , c. l. , woosley , s. e. , & heger , a. 2001 , apj , 550 , 372 fuhrmann , k. , axer , m. , & gehren , t. 1993 , a&a , 271 , 451 gustafsson , b. , edvardsson , b. , eriksson , k. , jrgensen , u. g. , nordlund , .
, & plez , b. 2007 , in preparation hartwick , f. d. a. 1976 , apj , 209 , 418 heger , a. & woosley , s. e. 2002 , apj , 567 , 532 hill , v. , plez , b. , cayrel , r. , beers , t. c. , nordstr " om , b. , andersen , j. , spite , m. , spite , f. , barbuy , b. , bonifacio , p. , depagne , e. , franois , p. , & primas , f. 2002 , a&a , 387 , 560 houdashelt , m. l. , bell , r. a. , & sweigart , a. v. 2000 , aj , 119 , 1448 iben , i. jr .
1983 , memorie della societa astronomica italiana , 54 , 321 iwamoto , n. , umeda , h. , tominaga , n. , nomoto , k. , & maeda , k. 2005 , science , 309 , 451 johnson , j. j. , herwig , f. , beers , t. c. , & christlieb , n. 2007 , apj , in press , astro - ph/0608666 jrgensen , u. g. , larsson , m. , iwamae , a. , & yu , b. 1996 , a&a , 315 , 204 karlsson , t. 2006 , apj , 641 , l41 komiya , y. , suda , t. , minaguchi , h. , shigeyama , t. , aoki , w. , & fujimoto , m. 2007 , apj , in press , astro - ph/0610670 korn , a. j. , grundahl , f. , richard , o. , barklem , p. s. , mashonkina , l. , collet , r. , piskunov , n. , & gustafsson , b. 2006 , nature , 442 , 657 korn , a. j. , shi , j. , & gehren , t. 2003 , a&a , 407 , 691 kupka , f. g. , piskunov , n. , ryabchikova , t. a. , stempels , h. c. , & weiss , w. w. 1999 , a&as , 138 , 119 kupka , f. g. , ryabchikova , t. a. , piskunov , n. , stempels , h. c. , & weiss , w. w. 2000 , baltic astronomy , 9 , 590 kurucz , r. 2006 , http://kurucz.harvard.edu/linelists/linesmol limongi , m. , chieffi , a. , & bonifacio , p. 2003 , apj , 594 , l123 luque , j. & crosley , d. r. 1999 , lifbase version 1.5 , sri international report mp 99 - 009 , sri international , http://www.sri.com/cem/lifbase/lifbase.pdf mashonkina , l. , korn , a. j. , & przybilla , n. 2007 , a&a , 461 , 261 mcwilliam , a. , preston , g. w. , sneden , c. , & searle , l. 1995 , aj , 109 , 2757 meynet , g. , ekstrm , s. , & maeder , a. 2006 , a&a , 447 , 623 2007 , in preparation nomoto , k. , tominaga , n. , umeda , h. , kobayashi , c. , & maeda , k. 2006 , nuc.phys.a , 777 , 424 , astro - ph/0605725 norris , j. e. , beers , t. c. , & ryan , s. g. 2000 , apj , 540 , 456 norris , j. e. , ryan , s. g. , & beers , t. c. 1997 , apj , 489 , l169 . 2001 , apj , 561 , 1034 norris , j. e. , ryan , s. g. , beers , t. c. , aoki , w. , & ando , h. 2002 , apj , 569 , l107 plez , b. & cohen , j. g. 2005 , a&a , 434 , 1117 prantzos , n. 2003 , a&a , 404 , 211 ryan , s. g. & norris , j. e. 1991 , aj , 101 , 1865 ryan , s. g. , norris , j. e. , & beers , t. c. 1996 , apj , 471 , 254 ryan , s. g. , norris , j. e. , & bessell , m. s. 1991 , aj , 102 , 303 salvadori , s. , schneider , r. , & ferrara , a. 2007 , mnras , submitted , astro - ph/0611130 schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , apj , 500 , 525 searle , l. & sargent , w. l. w. 1972 , apj , 173 , 25 , m. j. , yan , y. , mihalas , d. , & pradhan , a. k. 1994 , mnras , 266 , 805 , m. f. , cutri , r. m. , stiening , r. , weinberg , m. d. , schneider , s. , carpenter , j. m. , beichman , c. , capps , r. , chester , t. , elias , j. , huchra , j. , liebert , j. , lonsdale , c. , monet , d. g. , price , s. , seitzer , p. , jarrett , t. , kirkpatrick , j. d. , gizis , j. e. , howard , e. , evans , t. , fowler , j. , fullmer , l. , hurt , r. , light , r. , kopan , e. l. , marsh , k. a. , mccallon , h. l. , tam , r. , van dyk , s. , & wheelock , s. 2006 , aj , 131 , 1163 , w. w. & gallagher , a. 1966 , phys . rev . , 145 , 26 suda , t. , aikawa , m. , machida , m. n. , fujimoto , m. y. , & iben , i. jr . 2004 , apj , 611 , 476 umeda , h. & nomoto , k. 2002 , apj , 565 , 385 . 2003 , nature , 422 , 871 . 2005 , apj , 619 , 427 wisotzki , l. , christlieb , n. , bade , n. , beckmann , v. , khler , t. , vanelle , c. , & reimers , d. 2000 , a&a , 358 , 77 woosley , s. e. & weaver , t. a. 1995 , apjs , 101 , 181 yoshii , y. 1981 , a&a , 97 , 280 lllrrrr & @xmath139 & @xmath140 & @xmath141 & synth & @xmath142 & @xmath143 + & @xmath144 & @xmath140 & @xmath145 & @xmath146 & @xmath147 & @xmath148 + & @xmath149 & @xmath140 & @xmath150 & @xmath151 & @xmath152 & @xmath153 + & @xmath154 & @xmath155 & @xmath156 & @xmath157 & @xmath158 & @xmath159 + & @xmath160 & @xmath155 & @xmath161 & @xmath162 & @xmath163 & @xmath164 + & @xmath165 & @xmath166 & @xmath167 & @xmath168 & @xmath169 & @xmath170 + & @xmath171 & @xmath155 & @xmath172 & @xmath173 & @xmath174 & @xmath175 + & @xmath176 & @xmath166 & @xmath177 & @xmath178 & @xmath179 & @xmath180 + & @xmath181 & @xmath182 & @xmath183 & @xmath184 & @xmath185 & @xmath186 + & @xmath187 & @xmath140 & @xmath188 & synth & @xmath189 & @xmath190 + & @xmath191 & @xmath192 & @xmath193 & synth & @xmath194 & @xmath195 + & @xmath196 & @xmath140 & @xmath197 & synth & @xmath198 & @xmath199 + & @xmath200 & @xmath201 & @xmath202 & @xmath203 & @xmath204 & @xmath205 + & @xmath206 & @xmath182 & @xmath207 & @xmath208 & @xmath209 & @xmath210 + & @xmath211 & @xmath140 & @xmath212 & @xmath213 & @xmath214 & @xmath215 + & @xmath216 & @xmath140 & @xmath217 & @xmath218 & @xmath219 & @xmath210 + & @xmath220 & @xmath221 & @xmath222 & @xmath162 & @xmath223 & @xmath224 + & @xmath225 & @xmath226 & @xmath227 & @xmath228 & @xmath229 & @xmath230 + & @xmath231 & @xmath232 & @xmath233 & @xmath208 & @xmath234 & @xmath235 + & @xmath236 & @xmath237 & @xmath238 & @xmath213 & @xmath239 & @xmath240 + & @xmath241 & @xmath242 & @xmath243 & @xmath244 & @xmath245 & @xmath246 + & @xmath247 & @xmath242 & @xmath248 & @xmath249 & @xmath245 & @xmath246 + & @xmath250 & @xmath140 & @xmath251 & @xmath252 & @xmath253 & @xmath254 + & @xmath255 & @xmath140 & @xmath256 & @xmath257 & @xmath258 & @xmath259 + & @xmath260 & @xmath140 & @xmath261 & @xmath262 & @xmath263 & @xmath264 + & @xmath265 & @xmath266 & @xmath267 & @xmath268 & @xmath269 & @xmath270 + & @xmath271 & @xmath272 & @xmath273 & @xmath157 & @xmath274 & @xmath275 + & @xmath276 & @xmath277 & @xmath278 & @xmath279 & @xmath280 & @xmath281 + & @xmath282 & @xmath283 & @xmath284 & @xmath285 & @xmath286 & @xmath287 + & @xmath288 & @xmath266 & @xmath289 & @xmath290 & @xmath291 & @xmath292 + & @xmath293 & @xmath272 & @xmath294 & @xmath295 & @xmath296 & @xmath297 + & @xmath298 & @xmath299 & @xmath300 & @xmath301 & @xmath302 & @xmath303 + & @xmath304 & @xmath299 & @xmath305 & @xmath306 & @xmath307 & @xmath169 + & @xmath308 & @xmath272 & @xmath309 & @xmath310 & @xmath311 & @xmath312 + & @xmath313 & @xmath266 & @xmath314 & @xmath315 & @xmath316 & @xmath317 + & @xmath318 & @xmath319 & @xmath320 & @xmath321 & @xmath322 & @xmath292 + & @xmath323 & @xmath324 & @xmath325 & @xmath326 & @xmath327 & @xmath328 + & @xmath329 & @xmath140 & @xmath330 & @xmath331 & @xmath332 & @xmath333 + & @xmath334 & @xmath335 & @xmath336 & @xmath337 & @xmath338 & @xmath307 + & @xmath339 & @xmath340 & @xmath341 & @xmath173 & @xmath342 & @xmath303 + & @xmath343 & @xmath344 & @xmath345 & @xmath346 & @xmath347 & @xmath348 + & @xmath349 & @xmath299 & @xmath350 & @xmath306 & @xmath269 & @xmath351 + & @xmath352 & @xmath272 & @xmath353 & @xmath249 & @xmath354 & @xmath355 + & @xmath356 & @xmath357 & @xmath358 & @xmath359 & @xmath332 & @xmath333 + & @xmath360 & @xmath140 & @xmath361 & @xmath362 & @xmath363 & @xmath159 + & @xmath364 & @xmath299 & @xmath365 & @xmath315 & @xmath274 & @xmath275 + & @xmath366 & @xmath266 & @xmath367 & @xmath368 & @xmath269 & @xmath270 + & @xmath369 & @xmath277 & @xmath370 & @xmath371 & @xmath351 & @xmath372 + & @xmath373 & @xmath374 & @xmath375 & @xmath376 & @xmath275 & @xmath180 + & @xmath377 & @xmath277 & @xmath378 & @xmath337 & @xmath275 & @xmath180 + & @xmath379 & @xmath340 & @xmath380 & @xmath381 & @xmath382 & @xmath383 + & @xmath384 & @xmath374 & @xmath385 & @xmath386 & @xmath387 & @xmath333 + & @xmath388 & @xmath283 & @xmath389 & @xmath178 & @xmath351 & @xmath281 + & @xmath390 & @xmath357 & @xmath391 & @xmath337 & @xmath158 & @xmath392 + & @xmath393 & @xmath374 & @xmath394 & @xmath395 & @xmath355 & @xmath392 + & @xmath396 & @xmath277 & @xmath397 & @xmath359 & @xmath398 & @xmath399 + & @xmath400 & @xmath340 & @xmath401 & @xmath218 & @xmath269 & @xmath280 + & @xmath402 & @xmath319 & @xmath403 & @xmath376 & @xmath302 & @xmath404 + & @xmath405 & @xmath340 & @xmath341 & @xmath268 & @xmath406 & @xmath328 + & @xmath407 & @xmath344 & @xmath408 & @xmath409 & @xmath410 & @xmath411 + & @xmath412 & @xmath344 & @xmath413 & @xmath213 & @xmath382 & @xmath383 + & @xmath414 & @xmath340 & @xmath415 & @xmath416 & @xmath417 & @xmath280 + & @xmath418 & @xmath319 & @xmath419 & @xmath420 & @xmath274 & @xmath270 + & @xmath421 & @xmath140 & @xmath422 & @xmath423 & @xmath158 & @xmath287 + & @xmath424 & @xmath340 & @xmath425 & @xmath426 & @xmath307 & @xmath351 + & @xmath427 & @xmath428 & @xmath429 & @xmath252 & @xmath382 & @xmath430 + & @xmath431 & @xmath319 & @xmath432 & @xmath173 & @xmath433 & @xmath434 + & @xmath435 & @xmath319 & @xmath436 & @xmath213 & @xmath280 & @xmath437 + & @xmath438 & @xmath344 & @xmath439 & @xmath440 & @xmath441 & @xmath303 + & @xmath442 & @xmath140 & @xmath443 & @xmath244 & @xmath312 & @xmath175 + & @xmath444 & @xmath319 & @xmath445 & @xmath295 & @xmath296 & @xmath446 + & @xmath447 & @xmath340 & @xmath448 & @xmath449 & @xmath327 & @xmath297 + & @xmath450 & @xmath344 & @xmath451 & @xmath452 & @xmath316 & @xmath169 + & @xmath453 & @xmath357 & @xmath454 & @xmath455 & @xmath297 & @xmath456 + & @xmath457 & @xmath458 & @xmath459 & @xmath460 & @xmath461 & @xmath462 + & @xmath463 & @xmath464 & @xmath465 & @xmath466 & @xmath467 & @xmath169 + & @xmath468 & @xmath324 & @xmath469 & @xmath157 & @xmath470 & @xmath471 + & @xmath472 & @xmath335 & @xmath473 & @xmath474 & @xmath411 & @xmath456 + & @xmath475 & @xmath266 & @xmath476 & @xmath203 & @xmath434 & @xmath477 + & @xmath478 & @xmath335 & @xmath479 & @xmath480 & @xmath434 & @xmath477 + & @xmath481 & @xmath272 & @xmath482 & @xmath184 & @xmath169 & @xmath483 + & @xmath484 & @xmath266 & @xmath485 & @xmath244 & @xmath363 & @xmath437 + & @xmath486 & @xmath299 & @xmath487 & @xmath488 & @xmath404 & @xmath489 + & @xmath490 & @xmath272 & @xmath491 & @xmath480 & @xmath462 & @xmath456 + & @xmath492 & @xmath299 & @xmath493 & @xmath310 & @xmath270 & @xmath483 + & @xmath494 & @xmath495 & @xmath496 & @xmath497 & @xmath498 & @xmath327 + & @xmath499 & @xmath495 & @xmath500 & synth & @xmath383 & @xmath501 + & @xmath502 & @xmath503 & @xmath504 & @xmath505 & @xmath506 & @xmath507 + & @xmath508 & @xmath72 & @xmath509 & @xmath510 & @xmath511 & @xmath512 + & @xmath513 & @xmath503 & @xmath514 & @xmath515 & @xmath186 & @xmath516 + & @xmath517 & @xmath374 & @xmath518 & @xmath474 & @xmath519 & @xmath520 + & @xmath521 & @xmath374 & @xmath522 & @xmath420 & @xmath523 & @xmath524 + & @xmath525 & @xmath526 & @xmath527 & @xmath528 & @xmath519 & @xmath529 + & @xmath530 & @xmath526 & @xmath531 & @xmath249 & @xmath532 & @xmath533 + & @xmath534 & @xmath526 & @xmath535 & @xmath208 & @xmath536 & @xmath537 + & @xmath538 & @xmath140 & @xmath539 & @xmath262 & @xmath540 & @xmath541 + & @xmath542 & @xmath105 & @xmath543 & @xmath262 & @xmath544 & @xmath545 + & @xmath546 & @xmath547 & @xmath548 & @xmath262 & @xmath549 & @xmath550 + & @xmath551 & @xmath140 & @xmath552 & @xmath262 & @xmath553 & @xmath554 + & @xmath555 & @xmath140 & @xmath556 & @xmath262 & @xmath557 & @xmath558 + ll 53771.029 & 211.44 + 53771.072 & 211.67 + 53771.159 & 211.61 + 53791.025 & 211.98 + 53792.035 & 211.90 + 53792.078 & 212.02 + 53813.999 & 211.84 + llllll & 15.17 & 0.69 & 0.44 & 13.24 & 1.93 + he 0557@xmath14840 & 15.454 & 0.712 & 0.463 & 13.194 &
2.254 + lrlcrlrc @xmath559 & ... & 4900 @xmath52 100 & & ... & 5180 @xmath52 70 & @xmath1280 & 1 + @xmath37/j & 0.672 & 5170 @xmath52 50 & & 0.680 & 5150 @xmath52 50 & 20 & 2 + @xmath38/c & 0.438 & 5140 @xmath52 100 & & 0.434 & 5170 @xmath52 100 & @xmath130 & 2 + @xmath39/jg & 2.125 & 5060 @xmath52 70 & & 1.884 & 5340 @xmath52 80 & @xmath1280 & 2 + @xmath37/j & 0.672 & 5110 @xmath52 60 & & 0.680 & 5080 @xmath52 60 & 30 & 3 + @xmath39/jtcs & 2.179 & 4920 @xmath52 70 & & 1.935 & 5230 @xmath52 80 & @xmath1310 & 3 + lrrrr @xmath2 ( k ) & @xmath76 & @xmath560 & @xmath55 & @xmath561 + @xmath3 ( cgs ) & @xmath562 & @xmath74 & @xmath562 & @xmath74 + @xmath563}$ ] & @xmath564 & @xmath565 & @xmath566 & @xmath565 + @xmath62 ( kms@xmath31 ) & @xmath567 & @xmath565 & @xmath567 & @xmath565 lrrrrrrrrl & 1 & @xmath142 & @xmath143 & & @xmath568 & & & & synthesis + c(ch ) & & @xmath569 & @xmath570 & @xmath105 & @xmath571 & 8.39 & @xmath572 & @xmath573 & a - x & b - x + n(cn ) & & @xmath574 & @xmath575 & @xmath105 & @xmath576 & 7.78 & @xmath577 & @xmath578 & b - x + n(nh ) & & @xmath579 & @xmath580 & @xmath581 & @xmath571 & 7.78 & @xmath582 & @xmath583 & a - x + & 1 & @xmath147 & @xmath148 & & & 8.66 & @xmath584 & @xmath585 & + & 1 & @xmath152 & @xmath153 & & @xmath586 & 6.17 & @xmath587 & @xmath588 & + & 5 & @xmath158 & @xmath483 & @xmath589 & @xmath586 & 7.53 & @xmath590 & @xmath591 & + & 1 & @xmath185 & @xmath186 & & @xmath586 & 6.37 & @xmath592 & @xmath593 & + & 1 & @xmath189 & @xmath190 & & @xmath568 & 6.31 & @xmath590 & @xmath594 & + & 2 & @xmath595 & @xmath596 & & @xmath568 & 6.31 & @xmath597 & @xmath598 & + & 4 & @xmath219 & @xmath599 & @xmath374 & @xmath586 & 3.05 & @xmath600 & @xmath601 & + & 5 & @xmath230 & @xmath602 & @xmath105 & @xmath603 & 4.90 & @xmath604 & @xmath594 & + & 2 & @xmath253 & @xmath605 & & @xmath606 & 5.64 & @xmath607 & @xmath608 & + & 1 & @xmath263 & @xmath609 & & @xmath610 & 5.39 & @xmath611 & @xmath612 & + & 60 & @xmath467 & @xmath355 & @xmath613 & @xmath568 & 7.45 & @xmath614 & @xmath615 & + & 2 & @xmath307 & @xmath348 & & @xmath616 & 7.45 & & & + & 2 & @xmath617 & @xmath230 & & @xmath606 & 4.92 & @xmath618 & @xmath619 & co - addition + & 8 & @xmath152 & @xmath620 & @xmath105 & @xmath606 & 6.23 & @xmath621 & @xmath622 & + & 1 & @xmath540 & @xmath541 & & @xmath568 & 2.92 & @xmath623 & @xmath624 & + & 1 & @xmath544 & @xmath545 & & & 2.21 & @xmath625 & @xmath626 & + & 1 & @xmath549 & @xmath550 & & & 2.59 & @xmath627 & @xmath577 & + & 1 & @xmath553 & @xmath554 & & @xmath606 & 2.17 & @xmath628 & @xmath629 & + & 1 & @xmath557 & @xmath558 & & @xmath630 & 0.52 & @xmath631 & @xmath632 & + lcccc ca i 4226 & + 0.26 & + 0.02 & + 0.10 & 2.19 + ca ii 3706 & + 0.17 & 0.06 & & 2.22 +
ca ii 3933 & 0.00 & 0.02 & & 2.24 +
in table [ tab : rejectedlines ] we list absorption lines that were excluded from the abundance analysis for various reasons . in the case of rejections based on visual inspection of a line in the vlt
/ uves spectrum , it was rejected when either n.c . or j.e.n . ( or both ) suspected it to be blended .
lllrl & @xmath633 & @xmath634 & @xmath635 & blended + & @xmath636 & @xmath637 & @xmath638 & blended + & @xmath639 & @xmath640 & @xmath641 & blended + & @xmath642 & @xmath643 & @xmath644 & blended + & @xmath645 & @xmath646 & @xmath647 & blended + & @xmath648 & @xmath649 & @xmath650 & blended + & @xmath651 & @xmath652 & @xmath653 & blended + & @xmath654 & @xmath655 & @xmath361 & blended + & @xmath656 & @xmath657 & @xmath658 & affected by balmer line + & @xmath659 & @xmath657 & @xmath660 & blended + & @xmath661 & @xmath634 & @xmath662 & low @xmath0 + & @xmath663 & @xmath664 & @xmath665 & low @xmath0 + & @xmath666 & @xmath667 & @xmath668 & low @xmath0 + & @xmath669 & @xmath667 & @xmath670 & low @xmath0 + & @xmath671 & @xmath672 & @xmath673 & low @xmath0 + & @xmath674 & @xmath664 & @xmath675 & low @xmath0 + & @xmath676 & @xmath677 & @xmath678 & low @xmath0 + & @xmath679 & @xmath680 & @xmath681 & low @xmath0 + & @xmath682 & @xmath683 & @xmath684 & low @xmath0 + & @xmath685 & @xmath686 & @xmath687 & low @xmath0 + & @xmath688 & @xmath689 & @xmath690 & low @xmath0 + & @xmath691 & @xmath680 & @xmath692 & low @xmath0 + & @xmath693 & @xmath683 & @xmath694 & low @xmath0 + & @xmath695 & @xmath696 & @xmath697 & low @xmath0 + & @xmath698 & @xmath683 & @xmath699 & low @xmath0 + & @xmath700 & @xmath686 & @xmath701 & low @xmath0 + & @xmath702 & @xmath680 & @xmath703 & low @xmath0 + & @xmath704 & @xmath664 & @xmath705 & affected by balmer line + & @xmath706 & @xmath680 & @xmath707 & affected by balmer line + & @xmath708 & @xmath683 & @xmath709 & blended + & @xmath710 & @xmath683 & @xmath711 & affected by balmer line + & @xmath712 & @xmath713 & @xmath518 & blended + & @xmath714 & @xmath715 & @xmath358 & blended with a ch line + & @xmath716 & @xmath717 & @xmath718 & blended with a ch line + & @xmath719 & @xmath720 & @xmath721 & blended + & @xmath722 & @xmath634 & @xmath723 & blended + & @xmath724 & @xmath725 & @xmath647 & blended + & @xmath726 & @xmath727 & @xmath436 & blended + & @xmath728 & @xmath725 & @xmath729 & blended + & @xmath730 & @xmath720 & @xmath731 & blended + & @xmath732 & @xmath725 & @xmath733 & blended + & @xmath734 & @xmath735 & @xmath736 & blended + & @xmath737 & @xmath738 & @xmath739 & blended + & @xmath740 & @xmath741 & @xmath742 & affected by balmer line + & @xmath743 & @xmath744 & @xmath745 & blended with a ch line + | we report the discovery and high - resolution , high @xmath0 , spectroscopic analysis of the ultra - metal - poor red giant he 0557@xmath14840 , which is the third most heavy - element deficient star currently known .
its atmospheric parameters are @xmath2 = 4900k , @xmath3 = 2.2 , and [ fe / h ] = 4.75 .
this brings the number of stars with [ fe / h ] @xmath4 4.0 to three , and the discovery of he 0557@xmath14840 suggests that the metallicity distribution function of the galactic halo does not have a gap " between [ fe / h ] = 4.0 , where several stars are known , and the two most metal - poor stars , at [ fe / h ] @xmath5 5.3 .
he 0557@xmath14840 is carbon rich [ c / fe ] = + 1.6 a property shared by all three objects with @xmath6 } < -4.0 $ ] , suggesting that the well - known increase of carbon relative to iron with decreasing [ fe / h ] reaches its logical conclusion ubiquitous carbon richness at lowest abundance .
we also present abundances ( nine ) and limits ( nine ) for a further 18 elements . for species having well - measured abundances or strong upper limits , he 0557@xmath14840 is normal " in comparison with the bulk of the stellar population at [ fe / h ] @xmath5 4.0 with the possible exception of co. we discuss the implications of these results for chemical enrichment at the earliest times , in the context of single ( `` mixing and fallback '' ) and two - component enrichment models . while neither offers a clear solution , the latter appears closer to the mark .
further data are required to determine the oxygen abundance and improve that of co , and hence more strongly constrain the origin of this object .
+ he 05574840 ultra metal - poor and carbon - rich |
the concept of symmetry plays an important role both in physics and mathematics .
symmetries are described by transformations of the system , which result in the same object after the transformation is carried out .
they are described mathematically by parameter groups of transformations .
their importance ranges from fundamental and theoretical aspects to concrete applications , having profound implications in the dynamical behavior of the systems , and in their basic qualitative properties .
another fundamental notion in physics and mathematics is the one of conservation law .
typical application of conservation laws in the calculus of variations and optimal control is to reduce the number of degrees of freedom , and thus reducing the problems to a lower dimension , facilitating the integration of the differential equations given by the necessary optimality conditions .
emmy noether was the first who proved , in 1918 , that the notions of symmetry and conservation law are connected : when a system exhibits a symmetry , then a conservation law can be obtained .
one of the most important and well known illustration of this deep and rich relation , is given by the conservation of energy in mechanics : the autonomous lagrangian @xmath0 , correspondent to a mechanical system of conservative points , is invariant under time - translations ( time - homogeneity symmetry ) , and to denote the partial derivative of function @xmath1 with respect to its @xmath2-th argument . ]
@xmath3 = 0 \end{gathered}\ ] ] follows from noether s theorem , , the total energy of a conservative closed system always remain constant in time , `` it can not be created or destroyed , but only transferred from one form into another '' .
expression is valid along all the euler - lagrange extremals @xmath4 of an autonomous problem of the calculus of variations .
the conservation law is known in the calculus of variations as the 2nd erdmann necessary condition ; in concrete applications , it gains different interpretations : conservation of energy in mechanics ; income - wealth law in economics ; first law of thermodynamics ; etc .
the literature on noether s theorem is vast , and many extensions of the classical results of emmy noether are now available for the more general setting of optimal control ( see @xcite and references therein ) .
here we remark that in all those results conservation laws always refer to problems with integer derivatives . nowadays
fractional differentiation plays an important role in various fields : physics ( classic and quantum mechanics , thermodynamics , etc ) , chemistry , biology , economics , engineering , signal and image processing , and control theory @xcite .
its origin goes back three centuries , when in 1695 lhopital and leibniz exchanged some letters about the mathematical meaning of @xmath5 for @xmath6 .
after that , many famous mathematicians , like j. fourier , n. h. abel , j. liouville , b. riemann , among others , contributed to the development of the fractional calculus @xcite .
the study of fractional problems of the calculus of variations and respective euler - lagrange type equations is a subject of current strong research .
f. riewe @xcite obtained a version of the euler - lagrange equations for problems of the calculus of variations with fractional derivatives , that combines the conservative and non - conservative cases . in 2002 o.
agrawal proved a formulation for variational problems with right and left fractional derivatives in the riemann - liouville sense @xcite .
then , these euler - lagrange equations were used by d. baleanu and t. avkar to investigate problems with lagrangians which are linear on the velocities @xcite . in @xcite
fractional problems of the calculus of variations with symmetric fractional derivatives are considered and correspondent euler - lagrange equations obtained , using both lagrangian and hamiltonian formalisms . in all the above mentioned studies , euler - lagrange equations depend on left and right fractional derivatives , even when the problem depend only on one type of them . in @xcite problems depending on symmetric derivatives
are considered for which euler - lagrange equations include only the derivatives that appear in the formulation of the problem . in @xcite
riemann - liouville fractional integral functionals , depending on a parameter @xmath7 but not on fractional - order derivatives of order @xmath7 , are introduced and respective fractional euler - lagrange type equations obtained .
more recently , the authors have used the results of @xcite to generalize the classical noether s theorem for the context of the fractional calculus of variations @xcite . differently from @xcite , where the lagrangian point of view is considered , here
we adopt an hamiltonian point of view .
fractional hamiltonian dynamics is a very recent subject but the list of publications has become already a long one due to many applications in mechanics and physics .
we extend the previous optimal control noether results of @xcite to the wider context of fractional optimal control ( theorem [ thm : mainresult : fda06 ] ) .
this is accomplished by means ( i ) of the fractional version of noether s theorem @xcite , ( ii ) and the lagrange multiplier rule @xcite . as a consequence of our main result
, it follows that the `` total energy '' ( the autonomous hamiltonian ) of a fractional system is not conserved : a new expression appears ( corollary [ cor : mainresult ] ) which also depends on the fractional - order of differentiation , the adjoint variable , and the fractional derivative of the state trajectory .
we briefly recall the definitions of right and left riemann - liouville fractional derivatives , as well as their main properties @xcite .
let @xmath8 be a continuous and integrable function in the interval @xmath9 $ ] .
for all @xmath10 $ ] , the left riemann - liouville fractional derivative @xmath11 , and the right riemann - liouville fractional derivative @xmath12 , of order @xmath7 , are defined in the following way : @xmath13 @xmath14 where @xmath15 , @xmath16 , and @xmath17 is the euler gamma function .
if @xmath7 is an integer , then from and one obtains the standard derivatives , that is , @xmath18 let @xmath8 and @xmath19 be two continuous functions on @xmath9 $ ] .
then , for all @xmath10 $ ] , the following properties hold : 1 . for @xmath20 ,
@xmath21 2 .
for @xmath22 , @xmath23 3 . for @xmath20 , @xmath24 ( fundamental property of the riemann - liouville fractional derivatives ) .
[ rem : der : const : nz ] in general , the fractional derivative of a constant is not equal to zero .
the fractional derivative of order @xmath20 of function @xmath25 , @xmath26 , is given by @xmath27 when one reads `` riemann - liouville fractional derivative '' in the literature , it is usually meant ( implicitly ) the left riemann - liouville fractional derivative . in physics
, @xmath28 often denotes the time - variable , and the right riemann - liouville fractional derivative of @xmath29 is interpreted as a future state of the process @xmath29 . for this reason ,
right derivatives are usually neglected in applications : the present state of a process does not depend on the results of the future development .
following @xcite , and differently from @xcite , in this work we focus on problems with left riemann - liouville fractional derivatives only .
this has the advantage of simplifying greatly the theory developed in @xcite , making possible the generalization of the results to the fractional optimal control setting .
we refer the interested reader in additional background on fractional theory , to the comprehensive book @xcite .
in @xcite a formulation of the euler - lagrange equations is given for problems of the calculus of variations with fractional derivatives .
let us consider the following fractional problem of the calculus of variations : to find function @xmath4 that minimizes the integral functional @xmath30 = \int_a^b l\left(t , q(t),{_ad_t^\alpha } q(t)\right ) dt \ , , \end{gathered}\ ] ] where the lagrangian @xmath31 \times \mathbb{r}^{n } \times \mathbb{r}^{n } \rightarrow \mathbb{r}$ ] is a @xmath32 function with respect to all its arguments , and @xmath33 . in the case
@xmath34 , problem is reduced to the classical problem @xmath35 = \int_a^b l\left(t , q(t),\dot{q}(t)\right ) dt \longrightarrow \min \ , .\ ] ] [ thm : fracteleq ] if @xmath36 is a minimizer of problem , then it satisfies the _ fractional euler - lagrange equations _ : @xmath37 the following definition is useful in order to introduce an appropriate concept of _ fractional conservation law_. we recall that the classical noetherian conservation laws are always a sum of products ( as assumed in ) and that the fractional rule for differentiation of a product , in the sense of riemann - liouville , is enough complex ( see @xcite ) . with respect to this ,
our operator @xmath38 is useful .
this operator was introduced in @xcite and we refer the reader to this reference for several illustrative examples and remarks . here
we just mention that the operator @xmath38 has resemblances with the classical poisson bracket ( remark [ rem : pp ] )
. given two functions @xmath8 and @xmath19 of class @xmath39 in the interval @xmath9 $ ] , we define the following operator : @xmath40.\ ] ] for @xmath34 , operator @xmath41 is reduced to @xmath42 the linearity of the operators @xmath43 and @xmath44 imply the linearity of the operator @xmath41 .
[ eq : fcl ] we say that @xmath45 is a _ fractional conservation law _ if and only if it is possible to write @xmath46 in the form of a sum of products , @xmath47 for some @xmath48 , and for each @xmath49 the pair @xmath50 and @xmath51 satisfy one of the following relations : @xmath52 or @xmath53 along all the fractional euler - lagrange extremals ( along all the solutions of the fractional euler - lagrange equations ) .
we then write @xmath54 .
[ rem : pp ] for @xmath34 and coincide , and @xmath55 is reduced to @xmath56 which is the standard meaning of _ conservation law _ , a function @xmath57 preserved along all the euler - lagrange extremals @xmath58 , @xmath10 $ ] , of the problem .
this implies that if @xmath59 is a solution to the classical hamilton - jacobi equations of motion , then @xmath60 defines a conservation law of the hamiltonian equations with hamiltonian @xmath61 if @xmath62 or @xmath63 , where @xmath64 denotes the canonical poisson bracket operator . in the more general fractional context
, the hamilton - jacobi equations were recently derived in @xcite .
[ def : invadf ] functional is said to be invariant under the one - parameter group of infinitesimal transformations @xmath65 if , and only if , @xmath66 for any subinterval @xmath67 \subseteq [ a , b]$ ] .
having in mind that condition is to be satisfied for any subinterval @xmath67 \subseteq [ a , b]$ ] , we can rid off the integral signs in .
this is done in the new definition [ def : inv : gt ] .
the next theorem provides an extension of the classical noether s theorem to fractional problems of the calculus of variations .
[ theo : tndf ] if functional is invariant under , then @xmath68 \tau(t , q )
\\ + \partial_{3 } l\left(t , q,{_ad_t^\alpha q}\right ) \cdot \xi(t , q)\end{gathered}\ ] ] is a fractional conservation law ( definition [ eq : fcl ] ) .
using theorem [ theo : tndf ] , we obtain here a noether s theorem for the fractional optimal control problems introduced in @xcite : @xmath69 = \int_a^b l\left(t , q(t),u(t)\right ) dt \longrightarrow \min \ , , \\ _
ad_t^\alpha q(t)=\varphi\left(t , q(t),u(t)\right ) \ , , \notag\end{gathered}\ ] ]
together with the initial condition @xmath70 . in problem , the lagrangian @xmath71\times
\mathbb{r}^{n } \times \mathbb{r}^{m } \rightarrow \mathbb{r}$ ] and the velocity vector @xmath72 \times \mathbb{r}^{n } \times \mathbb{r}^{m } \rightarrow\mathbb{r}^n$ ] are assumed to be @xmath39 functions with respect to all the arguments . in agreement with the calculus of variations
, we also assume that the admissible control functions take values on an open set of @xmath73 . a pair
@xmath74 satisfying the fractional control system @xmath75 of problem , @xmath10 $ ] , is called a _
process_. [ th : p ] if @xmath74 is an optimal process for problem , then there exists a co - vector function @xmath76 such that the following conditions hold : * the hamiltonian system @xmath77 * the stationary condition @xmath78 with the hamiltonian @xmath79 defined by @xmath80 in classical mechanics , the lagrange multiplier @xmath81 is called the _ generalized momentum_. in the language of optimal control ,
@xmath81 is known as the _ adjoint variable_. [ def : extpont ] any triplet @xmath82 satisfying the conditions of theorem [ th : p ] will be called a _
fractional pontryagin extremal_. for the fractional problem of the calculus of variations one has @xmath83 , and we obtain from theorem [ th : p ] that @xmath84 comparing the two expressions for @xmath85 , one arrives to the euler - lagrange differential equations : @xmath86 .
we define the notion of invariance for problem in terms of the hamiltonian , by introducing the augmented functional as in @xcite : @xmath87 \\ = \int_a^b \left[{\cal h}\left(t , q(t),u(t),p(t)\right)-p(t ) \cdot { _
q(t)\right]dt \ , , \end{gathered}\ ] ] where @xmath79 is given by .
theorem [ th : p ] is easily obtained applying the necessary optimality condition to problem .
[ def : inv : gt ] a fractional optimal control problem is said to be invariant under the @xmath88-parameter local group of transformations @xmath89 if , and only if , @xmath90 d\bar{t } \\
= \left[{\cal h}(t , q(t),u(t),p(t))-p(t ) \cdot { _
q(t)\right ] dt \ , .\end{gathered}\ ] ] [ thm : mainresult : fda06 ] if the fractional optimal control problem is invariant under , then @xmath91 \tau - p(t ) \cdot \xi\ ] ] is a fractional conservation law , that is , @xmath92 \tau - p(t ) \cdot
\xi \right\ } = 0\ ] ] along all the fractional pontryagin extremals . for @xmath34
, the fractional optimal control problem is reduced to the classical optimal control problem @xmath93 = \int_a^b l\left(t , q(t),u(t)\right ) dt \longrightarrow \min \ , , \\
\dot{q}(t)=\varphi\left(t , q(t),u(t)\right ) \ , , \end{gathered}\ ] ] and we obtain from theorem [ thm : mainresult : fda06 ] the optimal control version of noether s theorem @xcite : invariance under a one - parameter group of transformations imply that @xmath94 is constant along any pontryagin extremal ( one obtains from setting @xmath34 ) .
the fractional conservation law is obtained applying theorem [ theo : tndf ] to the augmented functional .
theorem [ thm : mainresult : fda06 ] provides a new interesting insight for the fractional autonomous variational problems .
let us consider the autonomous fractional optimal control problem , the situation when the lagrangian @xmath1 and the fractional velocity vector @xmath95 do not depend explicitly on time @xmath28 : @xmath96 = \int_a^b l\left(q(t),u(t)\right ) dt \longrightarrow \min \ , , \\ _
ad_t^\alpha q(t)=\varphi\left(q(t),u(t)\right ) \ , .
\end{gathered}\ ] ] [ cor : mainresult ] for the autonomous problem the following fractional conservation law holds : @xmath97 in the classical framework of optimal control theory one has @xmath34 and our operator @xmath98 coincides with @xmath99 .
we then get from the classical result : the hamiltonian @xmath79 is a preserved quantity along any pontryagin extremal of the problem .
the hamiltonian @xmath79 does not depend explicitly on time , and it is easy to check that is invariant under time - translations : invariance condition is satisfied with @xmath100 , @xmath101 , @xmath102 and @xmath103 . in fact , given that @xmath104 , holds trivially proving that @xmath105 : @xmath106 using the notation in , one has @xmath107 and @xmath108 .
conclusion follows from theorem [ thm : mainresult : fda06 ] .
we begin by illustrating our results with two lagrangians that do not depend explicitly on the time variable @xmath28 .
these two examples are borrowed from @xcite and @xcite , where the authors write down the respective fractional euler - lagrange equations . here
, we use our corollary [ cor : mainresult ] to obtain new fractional conservation laws .
we begin by considering a simple fractional problem of the calculus of variations ( see ( * ? ? ?
* example 1 ) and @xcite ) : @xmath35 = \frac{1}{2}\int_0 ^ 1 \left(_{0}d_{1}^{\alpha}q(t ) \right)^2dt
\longrightarrow \min \ , , \quad \alpha > \frac{1}{2 } \ , .\ ] ] equation takes the form @xmath109 we conclude from corollary [ cor : mainresult ] that @xmath110 is a fractional conservation law .
let us now consider the following fractional optimal control problem @xcite : @xmath111 = \frac{1}{2}\int_0 ^ 1 \left[q^2(t)+ u^2(t)\right ] dt \longrightarrow \min \ , , \\ _ { 0}d_{1}^{\alpha}q(t)=-q(t)+u(t ) \ , , \notag\end{gathered}\ ] ] under the initial condition @xmath112 .
the hamiltonian @xmath79 has the form @xmath113 from corollary [ cor : mainresult ] it follows that @xmath114 is a fractional conservation law . for @xmath34 , the fractional conservation laws and give conservation of energy .
finally , we give an example of an optimal control problem with three state variables and two controls ( @xmath115 , @xmath116 ) .
the problem is inspired in ( * ? ? ?
* example 2 ) .
we consider the following fractional optimal control problem : @xmath117 for @xmath34 the control system serves as model for the kinematics of a car and - reduces to example 2 of @xcite . from corollary [ cor : mainresult ]
one gets that @xmath118 is a fractional conservation law .
main difficulty of our approach is related with the computation of the invariance transformations . to illustrate this issue ,
let us consider problem with @xmath119 in the classical case , since @xmath36 does not appear both in @xmath1 and @xmath95 , such a problem is trivially invariant under translations on the variable @xmath36 , condition is verified for @xmath34 with @xmath120 , @xmath121 , @xmath122 and @xmath123 . in the fractional case
this is not in general true : we have @xmath124 , but condition is not satisfied since @xmath125 and the second term on the right - hand side is in general not equal to zero ( remark [ rem : der : const : nz ] ) .
the fractional euler - lagrange equations are a subject of strong current study because of its numerous applications . in @xcite
a fractional noether s theorem is proved .
the fractional hamiltonian perspective is a quite recent subject , being investigated in a serious of publications .
one can say , however , that the fractional variational theory is still in its childhood .
much remains to be done .
this is particularly true in the area of fractional optimal control where results are a rarity .
the main study of fractional optimal control problems seems to be @xcite , where the euler - lagrange equations for fractional optimal control problems ( theorem [ th : p ] ) are obtained , using the traditional approach of the lagrange multiplier rule .
here we use the lagrange multiplier technique to derive , from the results in @xcite , a new noether - type theorem for fractional optimal control systems .
main result generalizes the results of @xcite . as an application
, we have considered the fractional autonomous problem , proving that the hamiltonian defines a conservation law only in the integer case @xmath34 .
gf was supported by ipad ( instituto portugus de apoio ao desenvolvimento ) ; dt by ceoc ( centre for research on optimization and control ) through fct ( portuguese foundation for science and technology ) , cofinanced by the european community fund feder / poci 2010 .
we are grateful to professor tenreiro machado for drawing our attention to the _
2nd ifac workshop on fractional differentiation and its applications _
, 1921 july , 2006 , porto , portugal , and for encouraging us to write the present work .
inspiring discussions with jacky cresson at the universit de pau et des pays de ladour , france , are very acknowledged .
el - nabulsi , r.a . , torres , d.f.m .
: necessary optimality conditions for fractional action - like integrals of variational calculus with riemann - liouville derivatives of order @xmath126
methods appl .
_ * 30*(15 ) , 19311939 ( 2007 ) | using the recent formulation of noether s theorem for the problems of the calculus of variations with fractional derivatives , the lagrange multiplier technique , and the fractional euler - lagrange equations , we prove a noether - like theorem to the more general context of the fractional optimal control . as a corollary
, it follows that in the fractional case the autonomous hamiltonian does not define anymore a conservation law .
instead , it is proved that the fractional conservation law adds to the hamiltonian a new term which depends on the fractional - order of differentiation , the generalized momentum , and the fractional derivative of the state variable . |
the superluminal source grs 1915 + 105 displays a rich diversity of lightcurve morphology , power density spectrum ( pds ) , quasi - periodic oscillations ( qpos ) , phase lags , and coherence .
the lightcurve morphology varies from out - burst to intermittent out - burst , to quiescence ( e.g. belloni et al .
1997 , 2000 )
. the pds shape can be either a broken power law with a flat top or a simple power law ( morgan , remillard , & greiner 1997 ) .
the qpo fundamental frequency ranges from mhz to 67 hz ( e.g. morgan , remillard , & greiner 1997 ) , and some qpos were detected up to the third harmonic ( cui 1999 ) .
several studies have related the temporal behaviors with the spectral properties .
for example , using the standard disk blackbody and power law model , muno , morgan , & remillard ( 1999 ) found that the 0.5 10 hz qpo frequency increases as the inner - disk color temperature increases from 0.7 to 1.5 kev .
chen , swank , & taam ( 1997 ) showed that the hardness ratio between the energy bands 11 30.5 kev and 2 11 kev is a good indicator for the presence of two types of qpos .
when this hardness ratio is above 0.1 , grs 1915 + 105 exhibited narrow 0.5 - 6 hz qpos .
broad qpos or no qpo were observed when this hardness ratio is below 0.1 . according to markwardt , swank , & taam ( 1999 ) ,
the 1 15 hz qpos were present when the power - law component became hard and intense , but the qpo frequency is correlated with the parameters of the thermal component . in these studies ,
the 0.5 10 hz qpos were treated as a single type of qpos . however
, our studies of the phase lag behaviors of the 0.5 10 hz qpos indicate that there are three types of 0.5 10 hz qpo behaviors which can be classified according to the qpo frequency : 0.5 2.0 hz , 2.0 4.5 hz , and 4.5 10 hz
. complex phase lag ( or hard lag ) behaviors have been observed in the 3 12 hz qpos of xte j1550564 ( wijnands , homan , & van der klis 1999 ) and in the 66.8 mhz qpo of grs 1915 + 105 ( cui 1999 ) .
wijnands , homan , & van der klis ( 1999 ) found two types of 3 12 hz qpos in xte j1550564 .
the first type of qpos have a broad qpo peak with a qpo frequency of 6 hz .
the hard lags of this type of qpos are negative at both the fundamental and first harmonic frequencies .
the second type of qpos , which have narrow peaks and a fundamental peak frequency of 3 hz , switch signs for the hard lags at the different harmonic frequencies .
similarly , cui ( 1999 ) found that the hard lags of the 66.8 mhz qpo in grs 1915 + 105 alternate from negative to positive values as the frequency increases from the fundamental to higher harmonic frequencies . in this letter
, we show the different phase lag behaviors of 0.5 10 hz qpos in 20 _ rossi x - ray timing explorer ( rxte ) _ observations of grs 1915 + 105 .
we find that the phase lag behaviors are closely related to the qpo fundamental frequency and the spectral states . in section 2 ,
we describe how the data were analyzed . in section 3 , the analysis results are presented . in the last section ,
we discuss the implications of these results .
to examine the phase lag behaviors , we selected the @xmath0 observations published in morgan , remillard , & greiner ( 1997 ) that showed qpos in the frequency range of 0.5 10 hz .
we also picked several other observations in late 1996 and 1997 which cover a big dip in the @xmath1 all sky monitor ( asm ) light curve of grs 1915 + 105 .
a full list of the observations is given in table 1 .
the radio fluxes at the times of some of these observations are given in pooley & fender ( 1997 ) and fender et al .
( 1999 ) .
we extracted light curves using three energy bands , 3.3 5.8 kev , 5.8 13.1 kev , and 13.1 41.0 kev .
the channels in the energy range of 2.5 3.3 kev were not used , because they have been binned together with the channels below 2.5 kev in the archival high time resolution data .
the counts were summed into time bins of 7.8125 ms , and timing analyses were performed on intervals of 256 seconds . the timing analyses included calculations of power density spectra and cross - spectra . in order to attenuate noise , we averaged the power spectra and cross - spectra over all observation intervals . in displaying the power density spectra , we used the leahy normalization ( leahy et al .
1983 ) without subtracting white noise .
phase lags and coherence between signals at two different energy channels were obtained from the cross - spectra .
we used monte carlo simulations to estimate the phase lag errors from the cross spectra . in the simulations , we assumed that the values of the real and imaginary parts had normal distributions , and used the fact that the cross - spectrum is a linear function of detector count rates .
we classified the grs 1915 + 105 lightcurve profiles into two general states .
one is the `` quiescent '' state in which the count rates on the 1-s time resolution lightcurves vary by less than 20% of the average count rates and there are no spikes wider than 10 seconds .
the other is the `` out - burst '' or `` flaring '' state which has spikes wider than 10 seconds and count rates varying by more than 20% of the average value , but does not have large scale structures in the light curves .
belloni et al .
( 2000 ) classified the observations of grs1915 + 105 into 12 classes ( hereafter belloni classes ) , based on their count rate and hardness characteristicsi .
our `` quiescent '' state corresponds to belloni classes @xmath2 and @xmath3 , and the `` flaring '' state includes classes @xmath4 , @xmath5 , @xmath6 and @xmath7 .
we used a simple power law model plus interstellar absorption to fit the photon spectra in the energy range of 2.5 40 kev .
though this simple model does not give good fits to the data , the photon power - law index @xmath8 is a good indicator of the spectral hardness .
we defined three spectral states : soft state ( @xmath8 = 2.0 3.0 ) , medium soft state ( @xmath8 = 3.0 4.0 ) , and very soft state ( @xmath9 ) .
the results are summarized in table 1 .
the observations are listed in the increasing order of the qpo fundamental frequency .
three types of hard lag behaviors were found in the observations .
the hard lags at the fundamental and first harmonic frequencies are both positive when the fundamental frequency is low . when the fundamental frequency increases above 2 hz but below 4.5 hz , the hard lags become negative at the fundamental frequency but stay positive at the first harmonic frequency . when the fundamental frequency moves to even higher values , no significant first harmonic peak can be detected . while there is a smooth transition between these behaviors , to highlight their differences we describe the three types separately in the following sub - sections .
figure 1 shows an example of the 0.5 2 hz qpos and the hard lags between the two energy bands ( 3.3 5.8 kev and 13.1 41.0 kev ) . both the fundamental and first harmonic peaks are significantly detected .
the hard lags at both peaks are positive and consistent with the values at the surrounding continuum frequencies . among the four observations that detected the 0.5 - 2 hz qpos ,
the coherence between the two energy bands at the qpo fundamental frequency is 0.95 or 0.96 , which is also consistent with the overall distribution of the coherence .
the photon spectrum power - law index is in the range of 2.6 2.7 , which indicates that grs 1915 + 105 was in the soft state . as the qpo fundamental frequency increases , the hard lag at the fundamental frequency decreases while the the hard lag at the first harmonic frequency increases .
the x - ray flux in the range of 3.3 13.1 kev was essentially the same among the four observations , and the source was in the `` quiescent '' state .
the 2.0 4.5 hz qpos have very peculiar phase lag behaviors ( for example , see figure 2 ) .
the hard lags are negative at the qpo fundamental frequency but turn positive at the first harmonic frequency .
they significantly deviate from the overall trend in the hard lag distribution over frequencies .
the photon power - law index is in the range of 2.5 3.0 .
thus grs 1915 + 105 was in the soft and quiescent state .
though the hard lags at the fundamental frequency appear to generally become more negative at higher fundamental frequencies , this trend is not as well defined as that for the 0.5 2.0 hz qpos . among the nine observations that detected the 2 - 4.5 hz qpos , the total 3.3 13.1 kev x - ray flux varied by a factor of more than three , and had no apparent correlations with the qpo fundamental frequency .
however , spectral analyses by trudolyubov , churazov , & gilfanov ( 1999 ) showed that the fundamental frequency positively correlates with the flux of the soft blackbody component that was obtained by fitting the spectrum with the standard disk backbody plus an exponentially cutoff power law .
the coherence is generally high at the fundamental frequency ( above 0.85 ) except for the observation on july 5th , 1997 , which may be treated as a special case because the analysis was done on the quiescent segments of an intermittent out - burst state .
the coherences in the intermittent out - burst state were significantly lower than the coherences in both the quiescent and flaring states .
the implications of this need to be further investigated .
when the qpo frequency is above 4.5 hz , the first harmonic qpo peak can not be significantly detected , and we do not see a change in the sign of the hard lags between the fundamental and where the first harmonic frequencies would be ( for example , see figure 3 ) .
the hard lags at both frequencies are negative . the coherence at the fundamental frequency between the two energy bands ( 3.3 5.8 kev and 13.1 41.0 kev ) is lower than for the previous two types of qpos .
the source was in the medium soft state with a photon spectrum power law index of 3.1 3.4 . unlike the previous two types of qpo , 4.5 10 hz qpos can be present in both quiescent and out - bursting states .
the 3.3 13.1 kev x - ray flux is also very diverse .
we have three observations that did not detect 0.5 10.0 hz qpos .
one common feature among these observations is that the source was in the very soft state with a photon power - law index above 4.0 ( table 1 ) .
similar results were also reported by muno , morgan , & remillard ( 1999 ) . among the three observations
, one has a pca 3.3 13.1 kev count rate of 6889 counts / s , compared to 16800 and 24005 counts / s in the other two observations .
therefore , the source was not necessarily in the very high state even though the spectra were very soft .
through studying the hard lag behaviors of the 0.5 10 hz qpos , we have highlighted them into three types that are closely related to different temporal and spectral properties .
the first type of qpos ( 0.5 2 hz ) have positive hard lags at the fundamental and first harmonic frequencies and were observed in the quiescent soft state . the second type of qpos ( 2 4.5 hz )
were also detected in the quiescent soft state but have opposite signs for the hard lags at the fundamental and first harmonic frequencies .
the third type ( 4.5 - 10 hz ) , which showed up in medium soft quiescent / out - burst states , do not have significant harmonic peaks .
there is a general trend that the hard lags at the qpo fundamental frequency decrease from positive to negative values as the fundamental frequency moves from 0.66 hz to higher values .
we also found that the photon spectra become softer as the qpo fundamental frequency increases .
after we presented this work to the 195th aas meeting in atlanta on january 14th , 2000 , it was brought to our attention that a recent paper by reig et al .
( 2000 ) also reported such trends in grs 1915 + 105 .
the opposite signs of the hard lags at the qpo fundamental and first harmonic frequencies do not necessarily mean that the photon arrival times are different at different qpo harmonic frequencies .
for example , the fourier transform of a decaying oscillating signal @xmath10 is : @xmath11 the power density has two qpo peaks at @xmath12 and @xmath13 respectively . for @xmath14 ,
the two dominant terms are : @xmath15 so we have @xmath16 and @xmath17 .
therefore , @xmath18 , i.e. they have opposite signs for the phase term even though the oscillations have no time delays from each other .
the two additional minor terms make the conclusion less obvious , but we still can find appropriate @xmath19 and @xmath12 to make the phase lags have opposite signs .
therefore , it may be misleading to interpret the phase lag at a particular qpo frequency as simply the time delay between two oscillating signals . in the classic multi - color thermal disk models ,
the negative hard lags can be explained by assuming that perturbations propagate from the inner disk to the outer disk .
the perturbations on the inner edge may come from the central object or from the magnetic fields that are being sucked into the central object .
zhang , cui , & chen ( 1997 ) and cui , zhang , & chen ( 1998 ) suggested that grs 1915 + 105 contains a black hole rapidly spinning in the same direction as the accretion disk .
in such a prograde system , the last stable orbit is much smaller than the non - spinning system .
therefore , strong perturbations on the inner edge of the accretion disk are very likely .
there may also be perturbations propagating inwards from the outer edge of the disk , which generate positive hard lags .
when the disk inner edge is farther away from the central object and thus the qpo frequency is smaller , it is natural to assume that the perturbations on the inner edge is weaker .
therefore , the inward perturbations would dominate over the outward perturbations , and the hard lags are expected to be positive .
this could explain why the lowest 0.5 2 hz qpos have positive hard lags while the faster ones have negative hard lags .
taam , chen , & swank ( 1997 ) proposed that the out - bursts in grs 1915 + 105 are due to the ejection of the inner disk
. a direct prediction of this model is that the inner disk edge is farther away from the central object in the out - burst state than in the quiescent state .
our results may pose problems to this prediction .
the observation made on july 5th , 1997 , for example , had long out - burst and quiescent segments .
we therefore performed separate timing analyses on the out - burst and quiescent segments .
we found that the quiescent state has a lower qpo fundamental frequency than the out - burst state , and thus the disk inner edge should be closer to the central object in the out - burst state than in the quiescent state if we assume the qpo fundamental frequency is somehow related to the sound crossing time across the disk inner edge .
lllllllll @xmath20 & y & @xmath21 & @xmath22 & 0.95 & 2.6 & @xmath23 & q & 1996 - 07 - 26 + @xmath24 & y & @xmath25 & @xmath26 & 0.96 & 2.6 & @xmath27 & q & 1996 - 08 - 03 + @xmath28 & y & @xmath29 & @xmath30 & 0.95 & 2.7 & @xmath31 & q & 1996 - 07 - 19 + @xmath32 & y & @xmath33 & @xmath34 & 0.95 & 2.7 & @xmath35 & q & 1996 - 08 - 10 + @xmath36 & y & @xmath37 & @xmath38 & 0.95 & 2.9 & @xmath39 & q & 1996 - 07 - 16 + @xmath40 & y & @xmath41 & @xmath42 & 0.92 & 2.8 & @xmath43 & q & 1997 - 10 - 08 + @xmath44 & y & @xmath45 & @xmath46 & 0.88 & 2.6 & @xmath47 & q & 1997 - 02 - 22 + @xmath48 & y & @xmath49 & @xmath50 & 0.88 & 2.6 & @xmath51 & q & 1996 - 10 - 29 + @xmath52 & y & @xmath53 & @xmath54 & 0.19 & 2.6 & @xmath55 & q & 1997 - 07 - 05 + @xmath56 & y & @xmath57 & @xmath58 & 0.93 & 3.0 & @xmath59 & q & 1996 - 07 - 11 + @xmath60 & y & @xmath61 & @xmath62 & 0.92 & 3.0 & @xmath63 & q & 1996 - 07 - 14 + @xmath64 & y & @xmath65 & @xmath66 & @xmath67 & 2.6 & @xmath68 & q & 1996 - 12 - 24 + @xmath69 & y & @xmath70 & @xmath71 & 0.91 & 2.9 & @xmath72 & q & 1996 - 08 - 25 + @xmath73 & n & @xmath74 & n / a & 0.88 & 3.1 & @xmath75 & f & 1996 - 08 - 18 + @xmath76 & n & @xmath77 & n / a & 0.55 & 3.4 & @xmath78 & f & 1997 - 07 - 07 + @xmath79 & n & @xmath80 & n / a & 0.65 & 3.2 & @xmath81 & q & 1996 - 08 - 31 + @xmath82 & n & @xmath83 & n / a & 0.12 & 3.1 & @xmath84 & f & 1997 - 07 - 05 + @xmath85 & n & @xmath86 & n / a & 0.68 & 3.4 & @xmath87 & f & 1996 - 10 - 15 + no & n & n / a & n / a & n / a & 4.4 & @xmath88 & f & 1996 - 06 - 29 + no &
n & n / a & n / a & n / a & 4.0 & @xmath89 & q & 1997 - 07 - 20 + no & n & n / a & n / a & n / a & 4.0 & @xmath90 & f & 1997 - 12 - 22 + | through studying the hard lags between the soft ( 3.3 5.8 kev ) and hard ( 13.0 41.0 kev ) photons of the 0.5 10 hz qpos in grs 1915 + 105 , we have classified them into three types : 0.5 2.0 hz qpos , 2.0 4.5 hz qpos , and 4.5 10 hz qpos .
they are closely related to different temporal and spectral states .
the first type of qpos ( 0.5 2 hz ) have positive hard lags at both the qpo fundamental and first harmonic frequencies .
these qpos were observed in the quiescent soft state .
the second type of qpos ( 2 4.5 hz ) , which were also detected in the quiescent soft state , have opposite signs of hard lags at the qpo fundamental and first harmonic frequencies . the third type ( 4.5 10 hz ) , which showed up in medium soft quiescent / out - burst states , do not have significant higher harmonic peaks .
there is a smooth transition between these three types of qpo behaviors .
we did not detect 0.5 10 hz qpos in the very soft state .
we discuss some of the implications of these results . |
the catalog and atlas of cataclysmic variables ( edition 1 - @xcite and edition 2 - @xcite ) has been a valuable source of information for the cataclysmic variable ( cv ) community .
one of the goals of the catalog was to have the basic information on the objects ( i.e. coordinates , type , magnitude range , and finding charts ) in one central location , thus making it easy for observers to obtain data on the objects .
however , the impracticality of reprinting the finding charts in their entirety means that , with each new edition , they are spread among more publications , taking us further from our goal of a central location .
furthermore , as new objects are discovered , and known ones examined in greater detail , the printed editions can not keep pace with discovery , a `` living '' edition is therefore highly desirable , so that observers can access a complete and current list of cvs at any time . for the above reasons , as well as
the need to simplify the tracking of the objects ( there are over 1200 objects in the catalog ) , we have decided to generate a web - based version of the catalog .
this version will have all the information ( as well some additional information detailed below ) from the first two editions , plus information on over 150 new objects discovered since 1996 may .
those objects with revised finding charts will only have one chart presented , thus eliminating a possible confusion which necessarily exists when `` paper '' catalogs are generated .
the web site will also allow for easy searching of the catalog , and for generation of basic statistics ( e.g. how many dwarf novae , how many cvs have _ hubble space telescope _
data , etc . ) .
the catalog consists of ( as of 2000 december ) 1034 cvs , and another 194 objects that are non - cvs ( objects originally classified erroneously as cvs ) .
most of the objects are dwarf novae ( 40% ) , with another 30% being novae , and the rest mostly novalike variables . a large fraction ( 90% ) of the cvs have references to published finding charts , while 64% of the objects have published spectra ( 49% quiescent spectra and 15% outburst spectra ) .
we have taken this opportunity to make several enhancements to the catalog . in conjunction with hans ritter and ulrich kolb
, we have added orbital period data to the catalog ; about one - third of the objects have periods .
the period information is from @xcite , plus updated and additional values . in conjunction with hilmar duerbeck @xcite , we now include finding charts of novae ( when possible ) , and have measured coordinates for many in the _ hubble space telescope _ gsc v1.1 guide star reference frame ( as is the case for the non - novae ) . finally , in the first edition we introduced ( out of necessity ) a pseudo - gcvs name for certain objects ( e.g. phe1 ) , which was continued in the second edition . with the web - based catalog ,
these names are no longer needed , so we will cease generating new ones .
for those objects that already had such names ( some of which have appeared in subsequent papers in the literature ) and now have a formal gcvs designation , we will adopt the formal gcvs name , although we will keep the pseudo - gcvs name in the `` other name '' field for continuity .
the site can be reached via : http://icarus.stsci.edu/@xmath0downes/cvcat/ and is described in detail below . the home page ( figure [ fig1 ] ) for the catalog contains six links : * * search * - a link to the search page , from which the catalog may be accessed . * * description * - a description of the catalog , following the format of the previous editions . a description of all the fields is given . * * references * - a complete listing of the references mentioned in the catalog .
note that from each individual object page , you can go directly to the reference of interest . * * statistics * - a listing of a fixed set of basic statistics from the catalog , generated in real - time . *
* ascii report * - a listing of the entire catalog in the format of the previously published versions ( i.e. containing most but not all of the fields ) , sorted by right ascension
. this output can be down - loaded to an ascii file . *
* change log * - a listing , by object , of the changes made since the initial release of this edition the search page ( figure [ fig2 ] ) is the main page for access to the catalog .
it allows the user to search the catalog on any field or combination of fields .
the following text fields can be searched in a case - insensitive manner : gcvs name , other name , and the five reference fields ( coordinate , chart , type , spectrum , and period ) ; the object type and notes fields can be searched in a case - sensitive manner .
all textual searches support the use of wildcards .
a coordinate search may be performed by specifying either a right ascension / declination range , or by specifying a set of coordinates and a radius .
numerical searches ( supporting a `` @xmath1 '' and `` @xmath2 '' capability ) can be performed for the following fields : galactic latitude , minimum and maximum magnitude , outburst year ( for novae ) , and period . finally , a search for space - based observations using any of 10 observatories can be performed .
an on - line help file is available detailing the search capabilities for each field , as well as providing instructions for the use of wildcards .
after a search is initiated , the search results page ( figure [ fig3 ] ) presents the results of the search .
this page indicates the number of objects in the catalog that match the selection criteria , and presents an abbreviated view of the catalog entries for such entries , showing the basic information such as the coordinates , type , magnitude range , and period . to obtain the full information ( including the finding chart ) , one clicks on the object of interest .
the individual object page ( figure [ fig4 ] ) presents the complete information on the selected object . for the finding charts , the field size , source ( dss ( digitized sky survey ) , _ hubble space telescope _
data , ground - based image ) , filter / emulsion , and exposure time are given to allow the user to estimate the depth of the image .
for most objects , the dss image is used .
however , for particularly crowded fields ( such as in globular clusters ) , _ hubble space telescope _
data is used when available .
similarly , for particularly faint targets , ground - based ccd images are provided when possible . on this page
, one may click on any of the reference codes to go directly to the full reference on the references page .
we plan to update the site with new objects and information on a continual basis , although period and spaced - based updates will occur roughly every six months .
we encourage users to inform us of any updates that should be implemented ( e.g. revised identifications , new objects , etc . ) , and if appropriate to send us improved / original finding charts ( as either postscript or jpeg images ) .
the charts will be particularly useful for recent novae recovered in quiescence , and for the faintest objects where deep ccd imaging clearly reveals the correct identification .
we wish to thank anne gonnella , steve hulbert , calvin tullos , and mike wiggs for the excellent work in creating the site .
we also wish to thank matt mcmaster for assistance in generating the multitude of finding charts , and the director s discretionary research fund at stsci for financial support .
paula szkody , john thorstensen , and steve howell provided helpful comments on the initial version of the site .
rfw gratefully acknowledges the support of nsf grant ast-9618462 , and sabbatical support from stsci .
hwd acknowledges the hospitality and support of stsci . | the catalog and atlas of cataclysmic variables ( edition 1 - @xcite and edition 2 - @xcite ) has been a valuable source of information for the cataclysmic variable ( cv ) community . however , the goal of having a central location for all objects is slowly being lost as each new edition is generated .
there can also be a long time delay between new information becoming available on an object and its publication in the catalog . to eliminate these concerns , as well as to make the catalog more accessible , we have created a web site which will contain a `` living '' edition of the catalog .
we have also added orbital period information , as well as finding charts for novae , to the catalog . |
metamaterials in the form of particulate composite mediums are currently of considerable scientific and technological interest walser . provided that wavelengths are sufficiently long compared with the length scales of inhomogeneities ,
such a metamaterial may be envisaged as a homogenized composite medium ( hcm ) , arising from two homogeneous component mediums l96 , mackay03 .
hcms with especially interesting properties may be conceptualized if the real parts of the relative permittivities ( and/or relative permeabilities ) of the two component mediums have opposite signs lijimmw .
this possibility arises for metal in insulator dielectric composites herwin , mlw01 and has recently become feasible with the fabrication of dielectric
magnetic materials displaying a negative index of refraction in the microwave frequency range helby , smith . over many years
, several theoretical formalisms have been developed in order to estimate the effective constitutive parameters of particulate composite mediums l96 .
in particular , the maxwell garnett and the bruggeman homogenization formalisms have been widely used ward .
generally , the maxwell garnett formalism is seen to hold only for dilute composite mediums mg .
more widely applicable is the bruggeman formalism that was initially founded on the intuition that the total polarization field is zero throughout the hcm brugge .
a rigorous basis for the bruggeman formalism is also available , within the framework of the strong permittivity fluctuation theory ( spft ) k81 , mlw00 .
estimates of hcm constitutive parameters generated by homogenization formalisms may be required to lie within certain bounds .
in particular , the wiener bounds wiener , aspnes and the hashin
shtrikman bounds s are often invoked . the hashin
shtrikman bounds coincide with the constitutive parameter estimates of the maxwell garnett homogenization formalism aspnes .
the applicability of theoretical bounds on the hcm permittivity has recently been the focus of attention for composites specified by relative permittivities with positive valued real parts ihvola . in this communication
, we consider the application of the bruggeman formalism , together with the wiener and hashin
shtrikman bounds , to isotropic dielectric hcms which arise from component mediums characterized by complex valued relative permittivities whose real parts have opposite signs .
this is scenario is typical of metal in insulator hcms aspnes , milton , for example . by duality
, our analysis extends to isotropic magnetic hcms .
it also extends to isotropic dielectric magnetic hcms , because the permeability and the permittivity are then independent of each other in the bruggeman formalism kampia ( as also in the maxwell garnett formalism lak92 ) .
therefore , our findings are very relevant to the application of homogenization formalisms lijimmw to mediums displaying negative index of refraction lmw03 , for example .
furthermore , the implications of our mathematical study extend beyond the bruggeman formalism to the spft as well mackay03 . a note on notation : an @xmath0 time
dependence is implicit in the following sections ; and the real and imaginary parts of complex valued quantities are denoted by @xmath1 and @xmath2 , respectively .
consider the homogenization of two isotropic dielectric component mediums labelled @xmath3 and @xmath4 .
let their relative permittivities be denoted by @xmath5 and @xmath6 , respectively . for later convenience , we define @xmath7 the bruggeman estimate of the hcm relative permittivity , namely @xmath8 , is provided implicitly via the relation ward @xmath9 wherein @xmath10 and @xmath11 are the respective volume fractions of component mediums @xmath3 and @xmath4 , and the particles of both component mediums are assumed to be spherical .
the bruggeman equation emerges naturally within the spft framework mackay03 .
a rearrangement of gives the quadratic equation @xmath12 only those @xmath8solutions of are valid under the principle of causality as encapsulated by the kramers kronig relations bh which conform to the restriction @xmath13 .
let @xmath14 be the discriminant of the quadratic equation ; i.e. , @xmath15 since @xmath16 , we may express as @xmath17 an insight into the applicability of the bruggeman formalism may be gained by considering the @xmath10roots of the equation @xmath18 ; these are as follows : @xmath19 on restricting attention to nondissipative component mediums ( i.e. , @xmath20 ) , it is clear that @xmath21 are complex valued if @xmath22 .
consequently , @xmath23 which implies that @xmath24 .
on the other hand , @xmath21 are real valued if @xmath25 .
thus , the bruggeman estimate @xmath8 for @xmath26 may be complex valued with nonzero imaginary part , even though neither component medium is dissipative .
various bounds on the hcm relative permittivity have been developed .
two of the most widely used are the wiener bounds wiener , aspnes @xmath27 and the hashin
shtrikman bounds s @xmath28 while both the wiener bounds and the hashin shtrikman bounds were originally derived for real valued constitutive parameters , generalizations to complex valued constitutive parameters have been established milton .
the hashin shtrikman bound @xmath29 is equivalent to the maxwell garnett estimate of the hcm relative permittivity @xmath30 based on spherical particles of component medium @xmath3 embedded in the host component medium @xmath4 .
similarly , @xmath31 is equivalent to the maxwell garnett estimate of the hcm relative permittivity @xmath32 based on spherical particles of component medium @xmath4 embedded in the host component medium @xmath3 .
the estimate @xmath30 is valid for @xmath33 , whereas the estimate @xmath32 is valid for @xmath34 ; but see the footnote in section 1 . to gain insights into the asymptotic behaviour of the wiener and hashin
shtrikman bounds , let us again restrict attention to the case of nondissipative component mediums ( i.e. , @xmath35 ) . from , we see that @xmath36 remains finite for all values of @xmath37 , but @xmath38 may become infinite for @xmath25 since @xmath39 in a similar vein , from we find that @xmath40 thus , for all values of @xmath41 there exists a value of @xmath42 at which @xmath43 is unbounded .
analogously , @xmath44 so we can always find a value of @xmath45 at which @xmath46 is unbounded , provided that @xmath47 .
let us now present , calculated values of the hcm relative permittivity @xmath8 , along with the corresponding values of the bounds @xmath48 and @xmath49 , for some representative examples .
both nondissipative and dissipative hcms are considered for @xmath50 .
the effects of dissipation may be very clearly appreciated through first considering the idealized situation wherein the components mediums are nondissipative vandehulst .
furthermore , although the absence of dissipation is unphysical due to the dictates of causality bh , weak dissipation in a particular spectral regime is definitely possible and is then often ignored ( * ? ? ?
* sec.2.5 ) .
thus , it is instructive to begin with the commonplace scenario wherein both @xmath51 and @xmath52 .
for example , let @xmath53 and @xmath54 . in figure 1
, @xmath8 is plotted against @xmath10 , along with the corresponding wiener bounds @xmath48 and hashin
shtrikman bounds @xmath49 .
the latter bounds are stricter than the former bounds in the sense that @xmath55 the close agreement between @xmath8 and the lower hashin shtrikman bound @xmath43 at low volume fractions @xmath10 is indicative of the fact that @xmath56 .
similarly , @xmath8 agrees closely with the upper hashin
shtrikman bound @xmath46 at high values of @xmath10 since @xmath57 .
a markedly different situation develops if the real valued @xmath58 and @xmath59 have opposite signs .
for example , the values of @xmath8 calculated for @xmath60 and @xmath54 are graphed against @xmath10 in figure 2 , together with the corresponding wiener and hashin shtrikman bounds .
the bruggeman estimate @xmath8 is complex valued with nonzero imaginary part for @xmath61 .
this estimate is not physically reasonable . the bruggeman homogenization formalism unlike the spft which is its natural generalization has no mechanism for taking coherent scattering losses into account .
furthermore , no account has been taken in the bruggeman equation for the finite size of the particles lijaem , pls , shan96 .
therefore , the bruggeman estimate of the hcm relative permittivity is required to be real valued if the component mediums are nondissipative . while @xmath8 in figure 2 is complex valued , the wiener bounds and the hashin shtrikman bounds are both real valued . in accordance with , we see that @xmath62 as @xmath63 .
similarly , @xmath64 in the limit @xmath65 , as may be anticipated from .
furthermore , since @xmath56 , the maxwell garnett formalism is clearly inappropriate here .
we also observe that the inequalities which hold for @xmath66 , do not hold for @xmath25 .
let us now investigate @xmath8 and its associated bounds when the component mediums are dissipative ; i.e. , @xmath67 .
we begin with those cases for which @xmath66 : for example , we take @xmath68 and @xmath69 . in figure 3
, @xmath8 is plotted against @xmath10 , and the associated wiener bounds @xmath48 and the hashin
shtrikman bounds @xmath70 are also presented .
the behaviour of the real parts of @xmath8 , @xmath48 and @xmath49 closely resembles that displayed in the nondissipative example of figure 1 .
in fact , the following generalization of holds : @xmath71 however , this ordering does not extend to the imaginary parts of @xmath8 , @xmath48 and @xmath49 . turning to the cases for @xmath25 , we let @xmath72 and @xmath69 , for example . the corresponding bruggeman estimate @xmath8 is graphed as function of @xmath10 , along with the wiener bounds @xmath48 and the hashin
shtrikman bounds @xmath49 in figure 4 . since @xmath73 ,
the real parts of @xmath38 and @xmath43 remain finite , unlike in the corresponding nondissipative scenario presented in figure 2 . however , the real and imaginary parts of @xmath38 and @xmath43 exhibit strong resonances in the vicinity of @xmath74 ( for @xmath38 ) and @xmath75 ( for @xmath43 ) .
these resonances become considerably more pronounced if the degree of dissipation exhibited by the component mediums is reduced . for example , in figure 5 the graphs corresponding to figure 4 are reproduced for @xmath76 and @xmath77 .
we observe in particular that @xmath78 for @xmath79 .
thus , the bruggeman estimate @xmath8 vastly exceeds both the wiener bounds @xmath48 and the hashin shtrikman bounds @xmath49 for a wide range of @xmath10 . since @xmath80 , the estimates of @xmath81 are clearly unreasonable .
furthermore , since the real and imaginary parts of @xmath82 exhibit sharp resonances at @xmath83 , we may infer that the maxwell garnett formalism is inapplicable for @xmath25 . on comparing figures 4 and 5 ,
we conclude that the bruggeman formalism , the weiner bounds and the hashin
shtrikman bounds become increasing inappropriate as the degree of dissipation decreases towards zero .
this means that all three _ could _ be applicable rather well when the dissipation is not weak .
therefore , let us examine the scenario wherein the real and imaginary parts of the relative permittivities of the component medium are of the same order of magnitude ; i.e. , we take @xmath84 and @xmath85 .
the corresponding plots of the bruggeman estimate @xmath8 together with the wiener bounds @xmath48 and the hashin
shtrikman bounds @xmath49 are presented in figure 6 .
the real and imaginary parts of the bruggeman estimate are physically plausible , and both lie within the hashin shtrikman bounds . the hashin
shtrikman bounds themselves do not exhibit resonances , and the weiner bounds do not exhibit strong resonances . accordingly , we conclude that many previously published results are not erroneous , but caution is still advised .
the bruggeman homogenization formalism is well established in the context of isotropic dielectric hcms , as well as more generally l96 . however
, this formalism was shown in section 3.2 to be inapplicable for hcms which arise from two isotropic dielectric component mediums , characterized by relative permittivities @xmath5 and @xmath6 , with * @xmath86 and @xmath87 having opposite signs ; and * @xmath88 @xmath89 @xmath90 .
since the bruggeman formalism provides the comparison medium which underpins the spft , it may be inferred that the spft is likewise not applicable to the scenarios of ( i ) with ( ii ) .
it is also demonstrated in section 3.2 that both the wiener bounds and the hashin
shtrikman bounds can exhibit strong resonances when the component mediums are characterized by ( i ) with ( ii ) . in the vicinity of resonances , these bounds clearly do not constitute tight bounds on the hcm relative permittivity . as a direct consequence , the maxwell garnett homogenization formalism , like the bruggeman homogenization formalism , is inapplicable to the scenarios of ( i ) with ( ii ) .
this limitation also extends to the recently developed incremental img and differential dmg variants of the maxwell garnett formalism .
if the component mediums are sufficiently dissipative then the bruggeman formalism and the hashin shtrikman bounds ( and therefore also the maxwell garnett formalism ) provide physically plausible estimates , despite the real parts of the component medium relative permittivities having opposite signs as shown in section 3.3 . the explicit delineation of the appropriate parameter range(s ) for the bruggeman formalism and the hashin
shtrikman bounds is a matter for future investigation .
bounds can , of course , be violated by a formalism if the underlying conditions for the formalism are in conflict with those used for deriving the bounds .
sihvola ihvola has catalogued the following conflicts : * bounds derived for nondissipative component mediums can be invalid for the real parts of either @xmath30 or @xmath32 for a composite medium containing dissipative component mediums .
* percolation can not be can not be captured by the maxwell garnett formalism herwin , sasta .
hence , the hashin shtrikman bounds , being based on the maxwell garnett formalism , can be violated by the bruggeman estimate @xmath8 for a percolative composite medium . *
the derivations of bounds generally assume that the particles in a composite medium have simple shapes .
if the particle shapes are complicated , the composite medium may display properties not characteristic of the either of the component mediums .
for instance , magnetic properties can be displayed when the particles in a composite medium have complex shapes @xcite , even though the component mediums are nonmagnetic .
clearly , the magnetic analogs of @xmath38 , @xmath36 , @xmath91 and @xmath92 are then inapplicable .
* @xmath38 , @xmath36 , @xmath91 , and @xmath92 as well as their magnetic analogs are also invalid _ prima facie _ when the component mediums exhibit magnetoelectric properties mackay03,lijaem , mich00 .
* bounds derived for electrically small particles become inapplicable with increasing frequency , due to the emergence of finite size effects pls .
even the concept of homogenization becomes questionable with increasing electrical size @xcite .
in contrast , the bounds and the homogenization formalisms studied in this paper share the same premises ; yet , a conflict arises in certain situations because the bounds exhibit resonance while the homogenization estimates do not .
as several conventional approaches to homogenization are not appropriate to the hcms arising from component mediums characterized by ( i ) with ( ii ) , there is a requirement for new theoretical techniques to treat this case .
this requirement is all the more pressing , given the growing scientific and technological importance of new types of metamaterials walser , lmw03 . | the bruggeman formalism provides an estimate of the effective permittivity of a particulate composite medium comprising two component mediums .
the bruggeman estimate is required to lie within the wiener bounds and the hashin shtrikman bounds .
considering the homogenization of weakly dissipative component mediums characterized by relative permittivities with real parts of opposite signs , we show that the bruggeman estimate may not be not physically reasonable when the component mediums are weakly dissipative ; furthermore , both the wiener bounds and the hashin shtrikman bounds exhibit strong resonances . * a limitation of the bruggeman formalism for homogenization * tom g. mackay + _ school of mathematics , university of edinburgh , edinburgh eh9 3jz , uk _ + akhlesh lakhtakia + _ catmas computational & theoretical materials sciences group + department of engineering science and mechanics + pennsylvania state university , university park , pa 168026812 , usa _
* keywords : * homogenization ; negative permittivity ; bruggeman formalism ; maxwell garnett formalism ; hashin shtrikman bounds ; wiener bounds |
it is well known that horizontal seismometers are sensitive to ground rotation at low frequencies.@xcite indeed , due to the equivalence principle , conventional seismometers and tiltmeters can not distinguish between horizontal acceleration and rotation of the ground .
while this is a problem in precision seismology , it is especially problematic for seismic isolation in next - generation gravitational - wave detectors , such as advanced laser interferometer gravitational - wave observatory ( aligo ) , where it is believed that rotation noise may limit improvements in low - frequency isolation.@xcite conventional horizontal seismometers can be idealized as horizontal spring - mass systems whose horizontal displacement is sensed relative to the housing . similarly , conventional tiltmeters or inclinometers can be idealized as vertical pendulums whose horizontal displacement is sensed relative to the housing .
they are schematically represented in fig .
[ tiltaccel ] . by `
tilt ' , we henceforth refer to the angular deflection of a simple pendulum with respect to its suspension platform or enclosure . from the diagram of the horizontal seismometer ,
a periodic rotation of @xmath5 at angular frequency @xmath6 will look equivalent to an acceleration of @xmath7 or a displacement of @xmath8 .
in other words , the displacement in response to a unit rotation for the horizontal seismometer is @xmath9 comparison between a horizontal seismometer , a tiltmeter , and our rotation sensor under the influence of slow horizontal acceleration and rotation . in the first two instruments
, an observer inside the box can not distinguish between the displacement of the bob due to rotation and acceleration . ]
due to the @xmath10 in the denominator , response to rotation dominates at low frequencies ( typically @xmath11 mhz ) .
similarly for a pendulum tiltmeter , the displacement of the bob due to a rotation is indistinguishable from that due to a horizontal acceleration at frequencies well below the resonance frequency .
consequently , the rotation in response to a unit displacement is given by the inverse of the right - hand side of eq .
( [ eq0 ] ) .
thus , typically , a tiltmeters output is dominated by acceleration at frequencies greater than @xmath12 mhz .
an important limitation of the active - control system for seismic isolation in aligo is this inability of horizontal seismometers to distinguish between ground rotation and horizontal acceleration at low frequencies ( @xmath13 to @xmath14 mhz@xcite ) .
slow ground rotation , such as that induced by wind , would be misinterpreted as large ground displacements , which could cause a large correction to be applied at the low - frequency active - isolation stages . this large displacement can produce noise in the gravitational - wave signal band through frequency up - conversion mechanisms , non - linearities , and cross - couplings .
this problem can be addressed by measuring the absolute ground rotation and removing it from the seismometer channels before it has a chance to enter the isolation system .
our rotation sensor design may also be important to the field of rotational seismology.@xcite large ring - laser gyroscopes are the traditional rotation sensors used in this field.@xcite they are traditionally run horizontally to sense rotation about the vertical axis , but can be mounted vertically to sense rotation about a horizontal axis.@xcite our design offers a simpler and more compact alternative . for comparison , it has roughly an order of magnitude better angle sensitivity than the horizontal - axis ring - laser gyroscope described in belfi _ _ et al.__@xcite between 10 to 100 mhz . in this frequency band ,
our sensor has comparable sensitivity to the angle sensitivity of c - ii : a vertical - axis meter - scale monolithic laser ring gyro@xcite and its sensitivity is surpassed by roughly an order of magnitude by the horizontal - axis 3.5 meter - square g-0 ring - laser gyro.@xcite
our rotation sensor is a low - frequency beam balance whose angle with respect to the platform is measured using an autocollimator . above the resonance frequency , the beam balance stays inertial as the platform rotates around it .
thus , the autocollimator measures the platform rotation , as shown in fig .
[ schematic ] . to decouple the influence of rotation and translation ,
the center of mass ( com ) of the balance is located as close to the axis of rotation as possible .
the relevant parameters of the balance are listed in table [ paratab ] .
schematic showing the principle of the rotation sensor . at frequencies above the resonance of the balance ,
it stays inertial as the platform rotates .
thus the autocollimator measures the platform rotation . ]
.[paratab]parameters of the balance [ cols="^,^",options="header " , ] to understand the dynamics of the system , we can write down the equations of motion in an inertial frame aligned with gravitational vertical
. let @xmath5 be the angle between the beam and the inertial frame s horizontal plane and let @xmath15 be the platform rotation angle with respect to the inertial frame s horizontal plane .
the equation of motion for rotation of the balance about the pivot axis in the presence of external torques ( assuming @xmath16 ) is @xmath17 where @xmath18 is the vertical distance from the com and the pivot ( positive sign if the com is below the pivot ) , @xmath19 is the stiffness of the flexure , @xmath20 is the total suspended mass of the balance , @xmath21 is its moment of inertia , @xmath22 is the loss factor of the flexure material ( @xmath23 ) , and @xmath24 is the horizontal displacement of the platform .
the external torque @xmath25 includes torques from all sources other than acceleration or rotation , such as brownian motion ( @xmath26 noise)@xcite , temperature gradients , or magnetic fields .
accelerations and rotations in other degrees of freedom are ignored in this simple model as they couple only through higher - order effects . to establish the transfer function for the autocollimator output in response to a platform rotation
, external torque and acceleration can be set to zero .
substituting @xmath27 and @xmath28 and solving for @xmath29 gives @xmath30 the quantity measured by the autocollimator is @xmath31 and the transfer function is @xmath32 the magnitude of the transfer function can be rearranged into the usual form for an oscillator as @xmath33 where @xmath34 , @xmath35 and @xmath36 .
the new quality factor @xmath37 can be either enhanced or reduced depending upon the sign of @xmath18 .
the transfer function has a zero at @xmath38 and a pole at @xmath39 . at frequencies much less than both @xmath38 and @xmath39 , @xmath40 fig .
[ transfunc ] shows the measured transfer function for a positive and a negative @xmath18 .
the low frequency response is given by eq .
( [ eq6 ] ) , which is strongly sensitive to @xmath18 . based on the sign of the gravitational spring stiffness @xmath41
there are three possibilities : 1 .
when @xmath41 is much larger than @xmath19 , the balance behaves as a simple pendulum , giving a low frequency response of @xmath42 .
when @xmath41 is close to zero , the balance behaves as a spring - mass system , giving a low frequency response much less than @xmath42 .
3 . when @xmath41 is @xmath43 , the low frequency response is enhanced .
this principle was exploited by building a second balance ( sec .
[ rotdata ] ) with a negative and large @xmath18 , which made it very sensitive to tilt .
magnitudes of the transfer function between platform rotation and autocollimator output for two configurations of the balance with different com . the measurement and a fit with free parameter @xmath18 are shown for each configuration . a positive @xmath18 results in a real zero in the transfer function whereas a negative @xmath18 yields an imaginary zero .
] thus , a simple way of minimizing @xmath44 is to minimize the low - frequency rotation response of the balance .
this measurement technique provides a method for minimizing @xmath44 far more precisely than would otherwise be possible .
similarly , we can calculate the transfer function for the autocollimator output in response to a platform acceleration .
setting external torque and platform rotation to zero gives @xmath45 rearranging terms as in eq .
( [ tiltresp2 ] ) yields @xmath46 and for @xmath47 , @xmath48 for our balance this is less than @xmath4 rad@xmath49 m .
this ratio is equivalent to that of a 33-km - long simple pendulum .
a useful way of understanding the noise sources on the balance is to calculate the net torque acting on the system .
the torque measured in the non - inertial lab frame can be written as @xmath50 using eqs .
( [ eq1 ] ) and ( [ autocoleq ] ) we can rewrite this as @xmath51 and converting to the frequency domain yields @xmath52 as mentioned previously , the first term includes torques from all non - acceleration sources . the second term in eq .
( [ eq9 ] ) contains the signal we are interested in measuring .
this can also be interpreted as a pseudo - torque .
the third term is the torque due to accelerations that couple through the gravitational spring .
drawing of the rotation sensor showing the beam balance suspended by a pair of flexures .
mirrors are located at the center for readout of the angular position of the balance . ]
a pair of beryllium - copper flexures are shown on the left .
the right side shows an image of the flexure as viewed through a microscope . ]
the balance beam consists of an aluminum tube with brass weights at the ends .
the tube has a diameter of 38 mm , wall thickness of 1.6 mm , and a length of 0.76 m. the brass weights are solid cylinders of diameter 34 mm and length of 0.23 m , and are fitted into the aluminum tube bringing the end to end length of the balance to 0.97 m. the balance is suspended by two flexures machined from beryllium - copper alloy , shown in figs . [ model ] and [ flex ] .
they are notch type flexures , with a thickness of @xmath53 @xmath3 m , radius of 3.2 mm , and width of 6.3 mm .
following standard procedures for achieving optimum tensile strength , the flexures were age - hardened at @xmath54 c for about 2 hours .
we built a vacuum chamber for the balance using stainless steel cf vacuum components ( see fig .
[ can ] ) .
the chamber consists of a 6-way cross with two tubes flanged to it .
a turbo vacuum pump backed by a diaphragm pump is connected through a long bellows to the 6-way cross .
the side - port contains a mechanical - manipulator bellows to adjust the horizontal com while the balance is under vacuum .
a viewport on top allows optical access for an autocollimator to measure the angle of the balance .
the instrument is located in a temperature - controlled ( @xmath55 c ) underground experimental hall and is surrounded by an extruded - foam enclosure .
the instrument is installed on a @xmath56-m - thick 600-kg aluminum plate bolted to three stainless steel feet .
the feet sit on a 0.3 m - thick concrete platform that rests on a deep foundation .
we installed 6.3-mm - thick aluminum thermal shielding tubes surrounding each balance - arm to provide thermal isolation and reduce thermal gradients .
these aluminum tubes are located inside the vacuum vessel and are thermally decoupled from the vacuum chamber and from the balance , but are connected to each other through thick aluminum heat links .
university of washington rotation sensor .
thermal insulation was removed for this picture . ]
the platform - angle - to - autocollimator - output transfer function was measured by applying a periodic horizontal force to the aluminum platform on which the balance sits .
this force deflected the platform rotationally ( @xmath57 @xmath3rad ) and induced horizontal translations ( @xmath57 @xmath3 m ) .
the periodic force was generated by a spring attached to a @xmath58-mm - diameter crankshaft driven by a stepper motor .
the spring was connected to the platform through a @xmath59 m cable which reduced the variation in the direction of the applied force .
the circular motion of the motor was thus converted to a linear force on the platform . due to the small @xmath18 during measurements , the response of the balance to displacements of the platform
were many orders of magnitude smaller than the rotation response .
hence the platform displacement did not affect the transfer function measurement .
each point in fig .
[ transfunc ] represents a single measurement which consisted of driving the platform at a single frequency for a few thousand seconds and measuring the autocollimator output .
schematic drawing of the four reflections of the light beam on the target mirror . ]
we used a multi - slit autocollimator , built by our group , that measures the angular deflection of a light beam reflected off a mirror mounted on the balance.@xcite the autocollimator uses a fiber - coupled led and a multi - slit grating as the light source and a sensitive ccd sensor for readout .
sensitivity was enhanced by employing multiple reflections of the light beam .
our setup used four reflections ( see fig .
[ fourb ] ) giving a dynamic range of 5 mrad .
the autocollimator is designed for use with a partially reflecting reference mirror in the beam path .
subtracting the reference pattern allows a rejection of common - mode noise such as low - frequency distortion of the vacuum can or the autocollimator body . in practice , we found such common - mode noise to be negligible compared with the seismic background .
however , this option allows referencing to a surface other than the platform to which the balance is bolted .
this would prove useful if we were to read the rotation of a surface that could not support the weight of the entire balance .
the reference pattern readout is also a measure of the intrinsic autocollimator noise .
the autocollimator was calibrated using one ordinary mirror and one partially reflecting mirror held at a fixed , known angle.@xcite in addition the instrument was calibrated using the gravitational torque from a pair of lead bricks on a turntable located @xmath60 m from the vacuum vessel .
the computed gravitational torque agreed with the measured torque to within the @xmath13 percent systematic uncertainty .
the rotation sensor output was recorded for several months . typically , the rotation noise was smallest from 22:00 hrs to 04:00 hrs .
the rotation in the 10-to 100-mhz frequency range was dominated by local weather , especially wind . to understand the spectral distribution ,
the raw time series of the autocollimator output is fourier transformed to an amplitude spectral density . to obtain the platform rotation
, we divide the autocollimator output by the measured transfer function between them ( eq . ( [ tiltresp ] ) ) .
[ tiltdata ] shows the spectral density of the data from a 4000-s measurement during a quiet period .
the spectral density shown is typical for most nights with low wind speeds .
typical platform rotation spectral density as measured by the rotation sensor during quiet conditions . also shown
are the autocollimator sensitivity and the rotation sensor requirement to potentially improve aligo.@xcite ] fig .
[ torque ] shows a plot of the torque spectral density as a function of frequency calculated using eq .
( [ torqeq ] ) for the same data set as above . also shown
are the thermal or brownian motion noise torque@xcite and the rotation sensor requirement,@xcite interpreted in torque units for our balance .
torque sensitivity of the rotation sensor . also shown
are the sensor requirements , autocollimator limit and thermal noise limit@xcite in torque units . ] separating the background rotation signal from the instrument noise requires a second instrument of similar or better sensitivity . since the rotation background at frequencies above 50 mhz was sufficiently small during quiet conditions , we focused on measuring the ground rotation near the 10-mhz limit of the required bandwidth . to build a sensitive tiltmeter at these frequencies we exploited the principle of the amplification of rotation using a negative @xmath18 , described in sec .
[ subsec : eqnmotion ] .
we constructed the tiltmeter from an aluminum plate and brass weights suspended by two stiff flexures ( fig .
[ newtilt ] ) .
this balance had its com far above the pivot ( @xmath61 mm ) resulting in a negative and large gravitational stiffness .
thus , an autocollimator mounted to the frame measured an amplified rotation signal at frequencies below resonance of the tiltmeter .
schematic of the tiltmeter .
it is suspended by two stiff flexures .
the com is located 78 mm above the pivot .
the angle of the beam is read - out using an autocollimator . ]
this tiltmeter was placed on the same aluminum platform as the rotation sensor . to check the calibration and sensitivity of this instrument , we temporarily modified the rotation sensor to run in tiltmeter mode by changing the com to be @xmath42 mm below the pivot .
the difference between the tilt measurements by the two sensors was smaller than @xmath62 nrad/@xmath63 at 10 mhz and less than @xmath42 nrad/@xmath63 at 30 mhz .
[ corr ] shows the angle spectral density from a 4000-s dataset recorded during the day with both instruments .
the angle measured by the rotation sensor and the ordinary tiltmeter agree well in the 15-to 60-mhz frequency range . above that , the angle measured by the tiltmeter is dominated by acceleration , which the rotation sensor rejects efficiently . fig .
[ msc ] shows the mean - squared coherence of the two data sets as a function of frequency .
subtracting the two datasets gives a factor of 2 to 3 suppression of rotation noise in the rotation sensor in the 15-to 60-mhz range .
the small @xmath18 of the rotation sensor balance implies that the microseismic acceleration affecting the tiltmeter in fig .
[ msc ] near @xmath64 hz would produce an angle noise of less than 10 picorad/@xmath63 .
this is confirmed by the coherence plot which shows poor coherence at the microseismic frequencies .
however , the residual noise after subtraction is worse compared to the noise from the quiet - condition data .
we believe this excess noise arises from high - frequency vibration , discussed in the next section . during quiet conditions at night , when background rotations were much smaller , the two instruments showed reduced correlation , indicating that at least one instrument was limited by intrinsic noise . .
there are various environmental influences that can produce noise in the rotation sensor . due to the low - pass filtering of torques by the balance and the nature of background influences ( such as weather ) ,
the noise requirement at 10 mhz is most challenging .
other than the autocollimator noise , all other noise sources are practically insignificant above 50 mhz .
comparison of the lowest torque noise measured with and without the rubber shims placed under the feet of the support platform , demonstrating the suppression of the torque noise . ] one of the largest noise sources we had to overcome was what we referred to as vibration noise .
the spectral shape of the noise was white ( in torque ) , but it had a daily amplitude variation of a factor of @xmath65 , correlated with ambient vibration levels .
there are several models which could explain this kind of noise .
one possible explanation is that this noise is a result of irregular torques arising from fluctuations of the parameters of the flexure ( parametric down - conversion ) triggered by excessive high - frequency vibrations .
regardless of the exact mechanism for the process , we found that placing small rubber shims under the feet of the large aluminum platform to which the rotation sensor was bolted reduced this vibration noise ( fig .
[ daily ] ) .
the white - noise floor was lowered by a factor of 10 with the introduction of the rubber shims .
still , this may remain the current limiting noise source and might be worse in places with large vibrations .
we plan to investigate and reduce this noise source further .
an important noise source is the presence of a temperature gradient between the two arms of the balance .
this armlength effect was analyzed by speake _
et al . _ @xcite using their formulation and a measured upper limit on the temperature gradient , we estimated the upper limit of the angle noise from this effect to be less than @xmath66 rad@xmath0 at 10 mhz .
noise from ambient magnetic field fluctuations exerting a torque on the balance are expected to be small .
magnetic materials were avoided in the construction of the balance .
the magnetic moment of the balance was measured to be @xmath67 @xmath68 j / t . with ambient field fluctuations expected to be no larger than @xmath69 t@xmath0
, the angle noise from magnetic field fluctuation would be less than @xmath70 rad@xmath0 .
another possible source of noise was from ambient gravity gradients .
our balance has a gravity gradient sensitivity of better than @xmath71 eotvos@xmath0 or @xmath72 s@xmath73 around @xmath13 mhz . for reference
, this would allow the rotation sensor to resolve the gradient signal from a 20-ton bus , with closest approach of 80 m at @xmath74 to the beam axis , in @xmath75 seconds to 2@xmath76 under quiet conditions . however , the expected gradient noise due to environmental sources such as nearby trees , buildings , etc .
, was estimated from models to be lower than our current best sensitivity by at least an order of magnitude .
sensitivity of the rotation sensor . ]
we have constructed a prototype absolute rotation sensor comprising of a @xmath77-mhz beam balance and an autocollimator .
we also constructed a tiltmeter or inclinometer to measure background rotation at frequencies below @xmath78 mhz .
the two instruments show good correlation in the @xmath79-to @xmath80-mhz band during the day when background rotations are larger . under quiet conditions
the correlation is small , indicating that the instruments are limited by intrinsic noise . using the best quiet - condition data
, we were able to place upper limits on the sensitivity of the instrument as shown in fig .
the sensor can reject horizontal acceleration to better than @xmath4 rad / m at frequencies above resonance . by design
, the rotation sensor has optimal sensitivity in a frequency band spanning @xmath81few mhz to hz .
the low - frequency range of the sensor is limited by the @xmath82 frequency - dependence in sensitivity below the @xmath77 mhz resonance frequency of the balance . above 0.1 hz
, the rotation sensor is limited by the @xmath64 nrad@xmath0 flat angular sensitivity of the autocollimator . at the high - frequency end ,
the sensor is limited by the 3.3 hz downsampling of the autocollimator signal.if desired , it can be operated at frequencies as high as 1.5 khz , excepting acoustic resonances of the balance .
this instrument has the potential to improve seismic isolation in advanced ligo by reducing the rotation noise contribution in horizontal seismometers .
it meets the requirements described by lantz _ _ et al.__@xcite above 40 mhz .
we continue to improve the sensitivity of the instrument at low frequencies and to develop a more robust and compact version .
we would like to thank ron musgrave , larry stark and jim greenwell at the uw physics instrument shop for our flexures .
we would like to thank the us taxpayers , nsf ( grants : phy0969488 and phy1306613 ) , and the ligo scientific collaboration for funding and supporting this project .
in particular , the authors would like to thank vladimir dergachev , brian lantz , fabrice matichard , and rainer weiss for valuable suggestions and discussions .
we are also grateful to our colleagues in the eot - wash group for useful discussions .
we thank the center for experimental nuclear physics and astrophysics ( cenpa ) for use of its facilities .
s. d. angelis and p. bodin , `` watching the wind : seismic data contamination at long periods due to atmospheric pressure - field - induced tilting '' , bull . of the seismological society of america ,
* 102 * , 1255 - 1265 ( 2012 ) b. lantz and r. schofield and b. oreilly and d. e. clark and d. debra , `` review : requirements for a ground rotation sensor to improve advanced ligo '' , bull . of the seismological society of america , * 99 * , 980 - 989 ( 2009 ) k. u. schreiber and j. n. hautmann and a. velikoseltsev and j.wassermann and h. igel and j. otero and f. vernon and j .- p .
r. wells , `` ring laser measurements of ground rotations for seismology '' , bull . of the seismological society of america ,
* 99 * , 1190 - 1198 ( 2009 ) j. belfi and n. beverini and f. bosi and g. carelli and a. d. virgilio and e. maccioni and a. ortolan and f. stefani , `` a 1.82 m@xmath83 ring laser gyroscope for nano - rotational motion sensing '' , e - print arxiv:1104.0418 ( 2011 ) t. b. arp and c. a. hagedorn and s. schlamminger and j. h. gundlach , `` a reference - beam autocollimator with nanoradian sensitivity from mhz to khz and dynamic range of @xmath84 '' , rev .
instrum . , * 84 * , 095007 ( 2013 ) c. c. speake and g. t. gillies , `` the beam balance as a detector in experimental gravitation '' , proceedings of the royal society of london .
series a , mathematical and physical sciences , * 414 * , 315 - 332 ( 1987 ) | we have developed a mechanical absolute - rotation sensor capable of resolving ground rotation angle of less than 1 nrad@xmath0 above @xmath1 mhz and 0.2 nrad@xmath0 above @xmath2 mhz about a single horizontal axis .
the device consists of a meter - scale beam balance , suspended by a pair of flexures , with a resonance frequency of 10.8 mhz .
the center of mass is located 3 @xmath3 m above the pivot , giving an excellent horizontal displacement rejection of better than @xmath4 rad / m .
the angle of the beam is read out optically using a high - sensitivity autocollimator .
we have also built a tiltmeter with better than 1 nrad@xmath0 sensitivity above 30 mhz .
co - located measurements using the two instruments allowed us to distinguish between background rotation signal at low frequencies and intrinsic instrument noise .
the rotation sensor is useful for rotational seismology and for rejecting background rotation signal from seismometers in experiments demanding high levels of seismic isolation , such as advanced ligo . |
nb / al - alo@xmath1/nb josephson junctions ( jjs ) with sizes in the sub-@xmath0 m to @xmath0 m range are required or advantageous for many applications such as high - speed superconducting digital circuits ( rsfq ) , mm - wave receivers and submillimeter wave mixers , programmable voltage standards , squids and superconducting qubits .
considering the multitude of possible applications , it is not surprising that several fabrication techniques for sub-@xmath0 m to @xmath0m - size junctions have been developed over the years . while a precise and reproducible definition of the jj size is routinely performed with the help of electron - beam lithography , the application of anodic oxidation in such processes remains a challenge .
this is due to the fact that the nb@xmath10o@xmath11 layer formed during the anodic oxidation creeps under the resist mask protecting the top electrode of the jj ( this is known as `` encroachment '' @xcite ) . even for a plasma hardened resist and an anodization voltage of 20 v ,
a significant encroachment was still observed @xcite . some groups solve this problem by simply omitting the anodic oxidation or performing it at very low voltages below 10 v @xcite . since we found that a certain nb@xmath10o@xmath11 thickness is needed to ensure low subgap leakage currents , this was not an option for us .
other groups have replaced the resist mask with an insulating hard mask such as sio or sio@xmath10 @xcite , which entirely eliminates the encroachment .
these hard masks have to be removed by an additional process step , either by reactive ion etching ( rie ) through the insulator right above the jj or by chemical mechanical polishing ( cmp ) . however , it has been found that plasma processes @xcite as well as cmp @xcite can damage the josephson junctions .
consequently , we have developed a novel method using an aluminum hard mask , which can be removed by a wet etching process with koh , which should not damage the josephson junctions .
additionally , wet etching is a very fast processing step , which helps to reduce the turnaround time .
this article is organized as follows : first , the newly developed process is discussed in some detail .
second , we present characteristic measurements , which show the high quality of our josephson junctions for all discussed applications .
finally , we present measurements on phase qubits and evaluate the experimentally observed coherence times , in order to see if our process is suitable for the fabrication of quantum devices .
] as a first step in the fabrication , nb / al - alo@xmath1/nb trilayers are deposited _ in - situ _ on an oxidized silicon substrate in a vacuum system containing nb and al magnetron sputtering targets .
both nb films are 100 nm thick and are dc sputtered at a power of @xmath12 w in the main chamber of our two - chamber vacuum system with a base pressure of around @xmath13 mbar .
special care has been taken to achieve stress - free nb films , which is crucial for high quality jjs @xcite .
this was done by systematically varying the ar working pressure @xmath14 for a nb film thickness of @xmath15 nm .
subsequently , the curvature of the silicon substrate bending under the film stress was measured , so that the latter could be determined quantitatively @xcite . for the optimal working point at @xmath16 mtorr , residual resistance ratios ( rrr ) of 3.1 and transition temperatures of @xmath17
k were obtained on 100 nm thick films , confirming the expected high quality of the nb electrodes .
the tunneling barrier is formed by dc sputtering of a 7 nm thick al layer at @xmath18 w and a subsequent oxidation of this layer in the load lock of the deposition system .
this oxidation is performed at room temperature in pure oxygen for @xmath19 min .
the critical current density of the josephson junctions can be reliably controlled in a wide range by the o@xmath10 pressure during oxidation ( see figure [ figure_trilayer_exposure ] ) , so that the required parameters can be achieved for a multitude of applications , ranging from superconducting qubits with @xmath20 a / cm@xmath2 to rsfq circuits at around 3 ka / cm@xmath2 .
afterwards , the shape of the bottom electrode ( including the alignment marks for the following electron - beam lithography ) is defined by positive photolithography . after this
, the trilayer is patterned by subsequent reactive - ion - etching ( cf@xmath22 + o@xmath10 ) of the top nb layer , ar ion beam etching ( ibe ) of the alo@xmath1 barrier and again rie of the bottom nb layer . in the following ,
the josephson junctions are defined with the use of a hard mask . as
already discussed , we wanted to avoid any further processes which might damage the junctions such as plasma etching or cmp .
hence , insulators like sio or sio@xmath10 were ruled out as hard mask material . to find an alternative ,
we searched for a metal that could be anodized , so that it would not act as a shortcut for the anodization current , and that could be removed fast and selectively by a clean wet etching process . since aluminum fulfills all of these criteria , the hard mask is created by positive electron - beam lithography and dc sputtering of a 50 nm thick al layer with subsequent lift - off .
it is then used for patterning of the junctions , which begins with etching the top electrode of the trilayer by rie . since aluminum is not etched here , both the alo@xmath1 tunneling barrier and the al mask act as ideal etch stoppers .
now , the first insulating layer is formed by anodic oxidation of the samples . here
, it is taken care that all sidewalls of the lower trilayer electrode are oxidized .
the anodization takes place in an aqueous solution of ( nh@xmath22)b@xmath11o@xmath23 and c@xmath10h@xmath24o@xmath10 at room temperature . per applied volt ,
0.88 nm nb are turned into 2.3 nm nb@xmath10o@xmath11 and 0.9 nm al are turned into 1.3 nm alo@xmath1 @xcite , so that for a typical anodization voltage of 25 v , only half the thickness of the al hard mask is oxidized and the underlying nb is protected .
the hard mask is removed by wet etching with koh , which is a fast process and should not damage the josephson junctions .
the jj surface after this processing step was investigated with scanning electron microscopy ( sem ) , as can be seen in figure [ figure_comp_masks ] . while encroachment of the nb@xmath10o@xmath11 was clearly visible on jjs which were defined using electron beam resist , we found indeed no encroachment for jjs defined by the al hard mask technique . ] ] after that , the second insulator sio is deposited employing positive e - beam lithography , thermal evaporation and lift - off . during evaporation , the wafers are water cooled in order to avoid thermally induced damage to the resist .
finally , the nb wiring layer is defined by negative photolithography and deposited in a magnetron sputtering system . here ,
an _ in - situ _ pre - cleaning by low - energy ar ibe is carried out before the sputtering , in order to provide a good electrical contact between the top junction electrode and the wiring layer .
this is especially required for jjs sizes in the sub-@xmath0 m range where the aspect ratios of the vias through the sio layer come close to one .
an overview of the entire fabrication process can be found in figure [ figure_fabr_subum ] .
over designed jj area @xmath25 for various circular junctions having a critical current density of @xmath26 a / cm@xmath2 , measured at 4.2 kelvin .
the fit shows an excellent agreement with the expected behavior @xmath27.[figure_jj_size_rn ] ] we have fabricated junctions of various critical current densities @xmath28 and characterized them at 4.2 kelvin .
for all applications discussed above , one of the most important quality parameters is the ratio of subgap resistance @xmath29 ( evaluated at a voltage of 2 mv ) over normal resistance @xmath30 . for the entire range of critical current densities shown in figure [ figure_trilayer_exposure ] ,
this @xmath31 ratio was typically larger than 30 , which indicates a very high quality . in the following
, the further discussion of junction quality will be carried out on two exemplary jj batches having different critical current density values . for a trilayer with @xmath32 a / cm@xmath2 , we systematically varied the jj size between 0.8 and 20 @xmath0m@xmath2 in order to see if the jj geometries were defined precisely .
since critical currents @xmath33 @xmath0a are strongly suppressed at @xmath34 k due to thermal noise , we took the normal resistance as the characteristic junction parameter .
figure [ figure_jj_size_rn ] shows the measured @xmath30 values versus the designed jj areas @xmath25 .
the data were fit by the power law @xmath35 .
it was found that @xmath36 , so that an excellent agreement with the expected behavior @xmath27 was obtained , which shows that the jj sizes are accurately defined by the al hard mask process .
this was confirmed by occasional measurements of the jj size with scanning electron microscopy ( sem ) .
furthermore , the fit value of @xmath37 @xmath38m@xmath2 leads to an average value of @xmath39 mv .
these high @xmath6 products show that we have strong cooper pair tunneling close to the expected ambegaokar - baratoff behavior @xcite .
four of these samples were cooled down to @xmath40 mk , so that measurements with unsuppressed critical currents were possible ; an exemplary @xmath3 curve can be seen in figure [ figure_iv_jj23 ] . here , we observed no excess currents and @xmath6 products of @xmath7 mv , which confirms the values given above .
furthermore , the gap voltages were found to be in the range of @xmath41 mv , which indicates a high quality of the nb electrodes .
the subgap branches were measured with a voltage bias and showed very low leakage currents , as can be seen in figure [ figure_iv_jj23 ] .
curve of a jj with @xmath26 a / cm@xmath2 and a diameter of @xmath42 @xmath0 m , measured at @xmath40 mk .
high quality parameters @xmath43 mv and @xmath44 mv are obtained . the subgap branch ( magnified by a factor of 200 )
shows low leakage currents ; it has been measured with a voltage bias and an applied magnetic field along the jj area , suppressing the critical current to 8 % of its zero - field value.[figure_iv_jj23 ] ] curve of a circular junction with an area of 0.5@xmath45m@xmath2 and a critical current density of @xmath46ka / cm@xmath2 , measured at 4.2 kelvin.[figure_iv_jc46_jj24 ] ] a different trilayer with @xmath47
ka / cm@xmath2 ( this @xmath28 value is of particular interest for the realization of rsfq circuits @xcite ) led to junctions with higher @xmath48 values . although the critical currents in the range of @xmath49 @xmath0a were still somewhat suppressed by thermal noise at @xmath34 k , we could now perform a full characterization at this temperature .
the recorded gap voltages accounted for @xmath50 mv and the characteristic voltages @xmath51 were all larger than 50 mv .
for jjs with areas @xmath52 @xmath0m@xmath2 , we observed values @xmath53mv , which we contribute to a less suppressed @xmath48 .
the @xmath3 curve of a sample with an area of 0.5 @xmath0m@xmath2 can be seen in figure [ figure_iv_jc46_jj24 ] .
since the @xmath31 ratios were very high for all critical current densities ( see above ) and high @xmath54 , @xmath6 and @xmath55 values were found at the exemplary critical current densities @xmath26 a / cm@xmath2 and @xmath47 ka / cm@xmath2 , we can conclude that a very high junction quality is obtained by the presented process .
furthermore , the junction size was accurately defined by the al hard mask technique .
m , @xmath56 m , @xmath57 m , @xmath58pf , @xmath59ph , @xmath60ph , @xmath61ph , @xmath62ph .
b ) photograph of the same structure with 1 = bias coil , 2 = qubit , 3 = squid , 4 = @xmath63 , 5 = mw line.[figure_phase_qubit_geometry ] ] high quality parameters extracted from the @xmath3 curves do not necessarily mean that the junctions can be successfully used for the fabrication of superconducting quantum bits , since decoherence mechanisms at higher frequencies , which are not detected in the @xmath3 curves , play an important role for such devices @xcite . consequently , to see
if the technological process described in this article is suitable for quantum bits , we fabricated phase qubits , of which one was characterized at mk temperatures in a dilution refrigerator .
we used a critical current density of @xmath64 a / cm@xmath2 and a standard phase qubit design similar to the one used in @xcite .
all vias necessary for the connection of the wiring to the lower electrode were not realized as large junctions , but the tunneling barrier was etched away with ibe , so that real vias with a direct connection between the nb bottom electrode and the nb wiring layer were formed .
schematics and a photograph of the circuit as well as design values can be found in figure [ figure_phase_qubit_geometry ] .
the design values lead to a characteristic qubit parameter of @xmath65 .
the setup and procedure of the measurement have been successfully used for the investigation of other phase qubits before and are described elsewhere @xcite .
the qubit is read out by application of a short ( 1 ns ) dc flux pulse , during which one magnetic flux quantum tunnels into the qubit loop only if the qubit was in the excited state @xmath66 .
the associated flux signal is detected by the inductively coupled squid , which results in two separated squid switching current histograms depending on whether the qubit was in the ground state @xmath67 or the excited state @xmath66 .
these switching histograms could be used to calculate the escape probability @xmath68 , which directly reflects the population of the qubit states .
first , by measuring the distance of these two separate squid switching peaks and taking into account the slope of the squid @xmath69 modulation , we determined the strength of the qubit flux signal in the squid to be @xmath70 m@xmath71 .
this is slightly higher than the expected design value of @xmath72 m@xmath71 .
such an increase in the qubit signal strength might be due to a higher critical current of the qubit @xmath73 than designed , which goes along well with the fact that also the observed average qubit level splitting @xmath74 ghz was slightly higher than expected . in the following , we chose a working point of @xmath75 ghz to characterize the qubit .
rabi oscillations of the system can be seen in figure [ figure_qubit_rabi_ramsey]a .
the oscillations decay exponentially with a characteristic time @xmath76ns .
furthermore , the relaxation time of the qubit was found to be @xmath8 ns . the dephasing time @xmath77 was determined by ramsey - type experiments with various detuning values of the applied microwave from the resonance frequency .
an example of such a measurement can be seen in figure [ figure_qubit_rabi_ramsey]b .
we determined the decay times of these exponentially decaying oscillations for various detuning values and took the mean value @xmath9 ns as our dephasing time .
details about each measurement procedure can be found in @xcite .
and @xmath66 at a level splitting of @xmath78ghz .
b ) ramsey oscillations for a detuning of 282 mhz from this working point.[figure_qubit_rabi_ramsey ] ] for most nb / al - alo@xmath1/nb phase qubits , the relaxation times were found to be between @xmath79 ns and @xmath80 ns @xcite .
for all these devices , the dephasing times @xmath77 could either not be measured or were shorter than the @xmath81 times .
this means that we have reached coherence times @xmath81 and @xmath77 which are certainly not shorter , but rather longer than for most other nb / al - alo@xmath1/nb phase qubits .
consequently , we can conclude that the novel al hard mask technique does not add any new sources of decoherence to the system and that the presented process is suited for the fabrication of quantum devices . in the future
, it would be interesting to measure the coherence times of qubits having identical design but being fabricated with and without the al hard mask technique , in order to see if this processing step has a systematic influence on decoherence .
we have developed a technological process for the fabrication of high quality sub-@xmath0 m to @xmath0m - size nb / al - alo@xmath1/nb josephson junctions .
this process employs an al hard mask , which acts as a perfect etch stop for rie and can be removed by a clean and fast wet etching step .
characterization of our junctions at 4.2 k and mk temperatures showed that we have high quality parameters @xmath82mv , @xmath83 mv at @xmath34 k and @xmath84 mv , @xmath85mv at @xmath40 mk . we fabricated phase qubits and characterized one of them at mk temperatures .
it exhibited a relaxation time of @xmath86ns and a dephasing time of @xmath87ns , which is comparable or even longer than for most other reported nb / alo@xmath1/nb phase qubits .
ketchen m. b , pearson d , kleinsasser a w , hu c - k , smyth m , logan j , stawiasz k , baran e , jaso m , ross t , petrillo k , manny m , basavaiah s , brodsky s , kaplan s b , gallagher w j and bhushan m 1991 _ appl .
lett . _ * 59 * 2609 feofanov a k , oboznov v a , bolginov v v , lisenfeld j , poletto s , ryazanov v v , rossolenko a n , khabipov m , balashov d , zorin a b , dmitriev p n , koshelets v p and ustinov a v 2010 _ nature physics _ * 6 * 593 | we have developed a combined photolithography and electron - beam lithography fabrication process for sub-@xmath0 m to @xmath0m - size nb / al - alo@xmath1/nb josephson junctions . in order to define the junction size and protect its top electrode during anodic oxidation
, we developed and used the new concept of an aluminum hard mask .
josephson junctions of sizes down to 0.5 @xmath0m@xmath2 have been fabricated and thoroughly characterized .
we found that they have a very high quality , which is witnessed by the @xmath3 curves with quality parameters @xmath4 mv and @xmath5 mv at 4.2 k , as well as @xmath6 products of @xmath7 mv obtained at lower temperatures . in order to test the usability of our fabrication process for superconducting quantum bits , we have also designed , fabricated and experimentally investigated phase qubits made of these junctions .
we found a relaxation time of @xmath8 ns and a dephasing time of @xmath9 ns . |
two - dimensional ( 2d ) materials have been extensively studied in the last decade @xcite owing to their applications in electronics and optoelectronics @xcite .
2d materials consist of layers that are held together by weak van der waals forces .
a remarkable feature of these layered materials is the difference in properties of a monolayer compared to multilayers of the same material @xcite .
for instance , monolayer of mos@xmath0 has a direct band gap , while multilayers of mos@xmath0 have an indirect gap @xcite .
most gapped 2d materials , like transition metal dichalcogenides ( tmdcs ) , hexagonal boron nitride and phosphorene , show an unmistakable reduction in band gap with the number of layers @xcite .
first principles electronic structure calculations , based on the gw @xcite approximation , have resulted in band gaps that are in excellent agreement with experiments @xcite on these materials .
band gaps of these materials calculated using density functional theory ( dft ) @xcite , while underestimated , also show a clear reduction with the number of layers @xcite .
most studies have attributed this reduction in the band gap to quantum confinement @xcite . however , there are no studies that _ quantitatively explain _ this trend .
the idea of quantum confinement comes from the model of an electron in a one - dimensional box . in multilayer stacks of 2d materials ,
the confining length of the box is directly proportional to the number of layers . in this context , the increase in the confinement length in multilayers as compared to a monolayer , due the summing of constituent layer potentials , is attributed to quantum confinement . on adding the potentials of constituent layers , the inter - layer spacing creates a finite barrier in the interstitial region between adjacent layers .
this barrier with a dft calculated hartree potential profile has been used to _ qualitatively _ explain the trend in the layer dependence of the band gap in phosphorene @xcite .
recently , perturbation approaches @xcite , which need input from explicit multilayer calculations , have been used to study layer dependence of band structures .
furthermore , in the case of tmdcs , it has been shown that the quasiparticle band gap calculated using the gw approximation converges extremely slowly with the number of unoccupied states , k - point sampling and the screened coulomb cut - off @xcite . performing separate gw calculations for a monolayer , bilayer , trilayer or more with parameters that ensure convergence is computationally very challenging . to study the variation of properties as a function of interlayer spacing or stacking of the layers needs one to repeat the calculation for different parameters .
moreover , gw calculations on heterostructures of 2d materials are presently intractable due to their lattice incommensurability necessitating the use of large supercells of each material .
identifying the origin of layer dependence opens up the possiblity of deriving ground state and excited state properties of multilayer stacks from calculations on a monolayer alone .
doing so would immensely ease the computation cost incurred in performing separate dft and gw calculations for different configurations of the constituent layers .
we study the physical origin of layer dependence in band structures of 2d materials .
we show that while quantum confinement gives _ qualitatively _ the right trend , the non - linearity of the @xmath1 functional plays a crucial role in _
quantitatively _ determining the layer dependence .
we show that within dft , band structures of multilayer stacks can be derived from a single calculation on a monolayer of the material .
we also extend this scheme to obtain quasiparticle energies of multilayer systems from a single gw calculation on a monolayer of the material .
we apply this method to understand the layer dependence of band structure in multilayers of a prototypical tmdc , 2h - mos@xmath0 , and compare the results to the standard calculations on this material @xcite .
we demonstrate the transferability of this scheme by applying it to graphene and phosphorene .
we performed the dft calculations using the plane - wave pseudopotential method as implemented in quantum espresso @xcite package .
norm - conserving pseudpotentials were used in all calculations .
the wave functions were expanded in plane - waves with an energy cut - off of 150 ry for mos@xmath0 . for graphene and phosphorene
, we used a wavefunction cut - off of 70 ry and 40 ry respectively .
the local density approximation to the exchange - correlation functional @xcite was used in mos@xmath0 and graphene calculations . for phosphorene we used the perdew - burke - ernzerhof exchange - correlation functional @xcite .
the brillouin zone was sampled with @xmath2 , @xmath3 and @xmath4 k - point grids for mos@xmath0 , graphene and phosphorene respectively .
we kept the in - plane lattice parameter for each material fixed in all the calculations .
the bilayer and trilayer were constructed from the monolayer with the appropriate stacking and inter - layer spacing .
no atomic relaxations were allowed in the bilayer and trilayer . the supercell dimension in the out - of - plane direction
was fixed at 35 .
the gw calculations were performed using the berkeleygw package @xcite . for mos@xmath0 , the dielectric function was evaluated with plane waves upto a cutoff of 35 ry and was extended to finite frequencies using the generalized plasmon pole ( gpp ) @xcite model .
the self - energy was constructed using the one shot @xmath5 method .
the coulomb interaction was truncated in the out - of - plane direction @xcite . for mos@xmath0
supercell size of 35 in the out - of - plane direction , @xmath2 k - point sampling and 8400 valence and conduction states are necessary to ensure convergence @xcite .
we perform separate calculations on monolayer , bilayer and trilayer mos@xmath0 with @xmath6 k - point sampling and 6000 valence and conduction states and compare the quasiparticle band structures of bilayer and trilayer with results obtained from the proposed scheme . these parameters ,
while not fully converged , demonstrate the efficacy of this scheme .
we also perform fully converged calculations on a monolayer of mos@xmath0 and derive the band gap , ionization potential and electron affinity of bilayer and trilayer using our scheme .
our fully converged monolayer band gap is in good agreement with previous calculations @xcite . for phosphorene , we perform separate calculations on monolayer , bilayer and trilayer with @xmath4 k - point sampling , 800 valence and conduction states , and 15 ry dielectric cutoff .
we compare the band gap , ionization potential and electron affinity obtained using our scheme to those obtained from the full calculation .
the static remainder technique was used to speed up convergence with the number of bands @xcite in all calculations .
( color online ) ( a ) bilayer mos@xmath0 in ab stacking , red spheres represent mo atoms and yellow spheres s atoms .
( b ) planar averaged charge density of bilayer mos@xmath0 ( black solid line ) , planar averaged charge density of layer a ( red dash - dot line ) , layer b ( blue dashed line ) and sum of layer a and b ( magenta line not seen ) .
the shaded regions indicate the position of the mos@xmath0 layers in the simulation cell .
( c ) planar averaged dft potential of the layer a , @xmath7 ( red dash - dot line ) , layer b , @xmath8 ( blue dashed line ) , the sum @xmath9 ( solid magenta line ) and bilayer , @xmath10 ( black solid line ) .
( d ) difference @xmath11 ( black solid line ) and @xmath12 - ( v_{xc}[\rho^a]+v_{xc}[\rho^b])$ ] ( violet dashed line ) . ] in order to understand the layer dependence of the dft band gap , consider a bilayer of mos@xmath0 with constituent layers labelled a and b ( fig .
[ fig1](a ) ) @xcite .
we perform separate dft calculations on layer a , layer b and the bilayer , to obtain the self - consistent charge densities and potentials of each . in fig .
[ fig1](b ) , we plot the planar averaged charge densities of layer a , layer b , the bilayer . from the figure
one can see that the bilayer charge density , @xmath13 , lies on top of the sum of the charge densities of the constituent layers , @xmath14+@xmath15 , which indicates that there is no significant rearragement of charge in the bilayer compared to the monolayers .
[ fig1](c ) shows the planar averaged total dft potential of the bilayer , @xmath10 . from this figure
, it is clear that the sum of the layer potentials , @xmath16 , is not equal to the bilayer potential .
the sum of the potentials , as described above , is the use of quantum confinement to obtain the potential for the bilayer .
the difference , @xmath17 , is localized to the interstitial region between the two layers where the charge densities overlap .
it can not arise from the hartree potential since it is a linear functional of the charge density and here the charge density of the bilayer is a sum of the charge densities of the individual layers . it can not arise from the ionic potential either , due to the absence of atomic relaxations in the bilayer . the difference must arise due to the non - linearity of the exchange - correlation functional . fig . [
fig1](d ) plots this difference which is an additional barrier between the layers . fig .
[ fig1](d ) also plots @xmath18 - ( v_{xc}[\rho^a]+v_{xc}[\rho^b ] ) = \delta v_{xc}$ ] , showing that this additional barrier indeed comes solely from the exchange - correlation difference .
the total bilayer potential can thus be expressed as a sum of the individual layer potentials and @xmath19 .
thus , the layer dependence of properties is an effect of quantum confinement and non - linearity of the @xmath1 functional .
( color online ) panels ( a ) , ( b ) and ( c ) plot the dft band structure of monolayer , bilayer and trilayer mos@xmath0 respectively .
panels ( d ) , ( e ) and ( f ) plot the dft band structures of monolayer , bilayer and trilayer graphene . the black solid line indicates the result from separate dft calculations on these systems .
the red dashed line shows the result obtained by a single - shot diagonalisation of the constructed hamiltonian eq [ eqn1 ] for these systems in the basis of the wavefunctions of the constituent layers .
the blue dashed lines are the eigenvalues obtained by considering only quantum confinement in these systems . ]
we can construct the dft hamiltonian for the bilayer in terms of the potential and charge density of the constituent layers as : @xmath20 where @xmath21 is the kinetic energy operator .
it should be noted that everything required to construct this hamiltonian can be obtained from a single monolayer calculation , say on layer a. based on the relative configuration of atoms in layer b with respect to layer a , a suitable transformation can be applied on @xmath22 and @xmath7 to obtain @xmath23 and @xmath8 respectively .
the wavefunctions of layer a , \{@xmath24 } , can similarly be transformed to obtain the wavefunctions of layer b , \{@xmath25}. the hamiltonian can then be constructed in the basis of the wavefunctions of the two layers , \{@xmath26 } , keeping in mind that the wavefunctions of the two layers do not form an orthogonal basis .
the generalized eigenvalue problem can be solved to yield eigenvalues and eigenfunctions of the bilayer .
this procedure can easily be generalized to n layers : @xmath27 ; where @xmath28 - \sum\limits_{i=1}^{n}v_{xc}[\rho^i]$ ] .
it is worth noting that while constructing the hamiltonian we use @xmath29 and not just the planar averaged quantities .
( color online ) modulus squared wavefunctions of monolayer mos@xmath0 , integrated out along [ 010 ] lattice vector direction .
z is the out - of - plane direction and x the [ 100 ] direction .
top panel : cbm wavefunctions at k , @xmath30 and @xmath31 points , respectively .
bottom panel : vbm wavefunctions at k , @xmath30 and @xmath31 points , respectively . ] the band structure of monolayer , bilayer and trilayer mos@xmath0 from separate dft calculations is plotted in fig .
[ fig2 ] ( a ) , ( b ) and ( c ) respectively . the points @xmath32 and @xmath30 are marked halfway between @xmath31-m and k-@xmath31 respectively .
[ fig2 ] ( d ) , ( e ) and ( f ) similarly show the band structure of monolayer , bilayer and trilayer graphene respectively . fig .
[ fig2 ] also shows the band structure obtained by neglecting @xmath19 from eqn .
( [ eqn1 ] ) .
this describes the effect of quantum confinement alone on the layer dependence of the band structures . while it captures the qualitative trends like the transition from direct to indirect band gap in mos@xmath0 , it fails to give an accurate layer dependence of the value of the band gap , ionization potentials and level splittings in the band structure . in mos@xmath0 ,
the splittings are overestimated at the conduction band minimum ( cbm ) of the @xmath30 point and underestimated at the @xmath31 point valence band maximum ( vbm ) .
the relative positions of the conduction band edges at the @xmath30 , k and @xmath32 points are also very different from the full dft calculation .
similarly , in graphene , excluding @xmath19 overestimates the ionization potential and the band splittings .
the band structure obtained from the eigenvalues obtained by diagonalizing the constructed hamiltonian described previously is also plotted in fig .
as can be seen , these band structures are in excellent agreement with the full calculation for mos@xmath0 and graphene .
a slight difference upto 5 - 10 mev is found due to a small rearrangement of charge in the bilayer or trilayer as compared to the sum of the constituent layer charge densities . hence to obtain the _ quantitative _ layer dependence , both the effects of quantum confinement and non - linearity of the exchange - correlation functional need to be accounted for .
the band structures of mos@xmath0 in fig .
[ fig2 ] show a transition from a direct band gap at k in the monolayer to an indirect band gap from @xmath31 to @xmath30 in the bilayer .
the transition is driven by the large splitting of the vbm at the @xmath31 point and cbm at the @xmath30 point . the k point vbm and cbm on the other hand split only slightly .
the amount by which a band splits depends on the off - diagonal elements of the multilayer hamiltonian represented in the basis of the constituent layer wavefunctions and the overlap between the wavefunctions of the constituent layers .
the vbm and cbm of the monolayer at k are localised in space , have a large mo d orbital character [ see fig [ fig3 ] ] .
the cbm at @xmath30 and vbm at @xmath31 on the other hand have a strong s p@xmath33 character [ see fig [ fig3 ] ] .
they are hence more delocalized in the out - of - plane direction and hybridize more with the wavefunctions of other layers than the wavefunctions at k. this leads to the large splittings in the band structure at these points for the bilayer and trilayer [ see fig [ fig2 ] ( b ) and ( c ) ] . the vbm and cbm at k , vbm at @xmath30 and cbm at @xmath31 show little to no spliting in the band structures owing to their localized nature
[ see fig [ fig3 ] ] .
we can extend this method to calculate the quasiparticle energies and band structures .
the dft eigenvalues can be corrected by using the gw approximation to the electron self - energy , @xmath32 @xcite . for the bilayer ,
quasiparticle eigenvalues are given by : @xmath34 where @xmath35 is the dft wavefunction corresponding to the eigenvalue @xmath36 and @xmath37 is the corresponding quasiparticle energy .
evaluating @xmath38 with the one - shot @xmath39 method within the gpp approximation @xcite requires the bilayer charge density , the bilayer irreducible polarizability , @xmath40 , and wave functions of the bilayer , @xmath41 .
we now show that all the required quantities can be approximated from the quantities obtained from a monolayer calculation , say on layer a. the procedure to obtain @xmath41 and @xmath13 is as described before .
the bilayer self - energy can be written as a sum over the individual self - energies of the layers and a correction term : @xmath42 @xmath43 can be computed directly from monolayer irreducible polarizability , @xmath44 , @xmath22 and @xmath45 . to compute @xmath46
, we obtain @xmath47 , @xmath23 and @xmath48 by applying transformations similar to the ones described above to @xmath44 , @xmath22 and @xmath45 respectively . the correction term
, @xmath49 contains information on the interaction between the layers . due to the weak coupling between the layers ,
we expect @xmath49 to be a small correction ( compared to @xmath50 ) .
we can evaluate @xmath51 at various levels of approximation .
it can be evaluated at the dft level by @xmath52 ; or assuming just exchange interaction between the layers @xmath53 ; or within the static limit of gw ( cohsex ) @xmath54 ; or within full gw .
the only additional quantity needed in some of these approximations is , @xmath40 , which can be approximated as a sum of the irreducible polarizability of constituent layers : @xmath55 @xcite .
this method can easily be extended to calculate band structures of n layers by computing @xmath56 .
@xmath57 can then be evaluated at an appropriate level of approximation . .
ionization potential ( ip ) , electron affinity ( ea ) and band gap ( in ev ) of bilayer and trilayer mos@xmath0 evaluated using the constructed @xmath32 for various approximations of @xmath51 described in the text .
the top section compares the results from full calculations performed with 12@xmath5812@xmath581 sampling , 6000 bands to the results from various approximations of @xmath51 .
@xmath59 denotes @xmath51 evaluated at the gpp level with 800 bands .
the bottom section are the results obtained by applying our scheme to a fully converged monolayer calculation .
[ cols="^ , > , > , > , > , > , > " , ]
( color online ) ( a ) , ( b ) and ( c ) dft charge density , potential and band structure of bilayer mos@xmath0 at the equilibrium interlayer spacing of 6.0 , respectively .
( d ) , ( e ) and ( f ) for interlayer spacing of 7.5 .
( g ) , ( h ) and ( i ) for interlayer spacing of 9.0 . ]
( color online ) ( a ) band gap of bilayer mos@xmath0 as a function of interlayer spacing evaluated using the constructed @xmath32 for various approximations to @xmath60 .
the dashed line marks the monolayer band gap within gw .
( b ) vbm and cbm of the bilayer as a function of interlayer spacing , evaluated using @xmath61 .
the dashed line marks the gw vbm and cbm of a monolayer . ]
the transition in bilayer mos@xmath0 , from interacting monolayers to non - interacting ones can be studied by gradually increasing the interlayer spacing .
[ fig5 ] shows the evolution of the dft band structure , charge density and potential of bilayer mos@xmath0 as a function of increasing interlayer spacing , d. the equilibrium spacing is @xmath62 . as the spacing between the layers increase , the interaction between them weakens and the splitting of the valence bands at @xmath31 and the conduction bands at @xmath30 reduces
this leads to a band gap transition from @xmath63 to @xmath64 to @xmath65 . at d = 7.5 @xmath66 ,
the charge densities of the two layers stop overlapping , but the potential barrier between the layers is not zero . at this point
the layers are weakly interacting and the nature of interaction within dft is purely due to quantum confinement ; @xmath19 is zero . at d = 9 @xmath66 and above , the two layers are completely non - interacting at the dft level and the gap is that of monolayer mos@xmath0 .
we construct the bilayer self energy at different interlayer spacings using the various approximations to @xmath60 .
the gap thus computed is shown in fig .
[ fig6 ] ( a ) . in the approximation of @xmath67 and @xmath68 ,
the layers become non interacting once the charge densities of the two layers stop overlapping .
thus the gap in this approximation goes to the monolayer gap for spacing larger than d = 7.5 .
the approximations of @xmath69 and @xmath61 include the long range correlation interaction between the layers .
the band gap computed in these approximations thus show a slower convergence to the monolayer gap with increasing interlayer spacing .
note that these calculations are performed using a coarser 12@xmath5812@xmath581 k - point sampling and 6000 bands , leading to an overestimate of the monolayer gap in fig .
[ fig6 ] ( b ) shows the vbm and cbm levels computed using @xmath61 , as a function of increasing interlayer spacing .
the bilayer cbm shows a weak dependence on the interlayer spacing and is already aligned with the monolayer cbm , while the bilayer vbm shows a slow convergence towards the monolayer vbm as the spacing is increased .
this is similar to the effect of a metallic substrate on a molecule , where dft shows no renormalization of the band gap but accounting for correlation effects in gw shows a significant renormalization @xcite .
the renormalization of the molecular levels is due to screening from image charge effects , arising from the metal substrate . in the bilayer mos@xmath0 system ,
similarly , one monolayer feels the screening from the other .
we studied the origin of layer dependence in band structures of 2d layered materials and developed a scheme to derive multilayer properties from calculations on a monolayer alone .
we showed that the observed trend in layer dependence within dft is a combined effect of quantum confinement and non - linearity of the dft exchange - correlation functional .
we also constructed the electron self energy for multilayers in terms of monolayer irreducible polarizability , charge density and wavefunctions .
the dft and quasiparticle band structures obtained using this scheme are in excellent agreement with those from the full calculation .
the advantage of this scheme is that it can provide accurate results operating at a small fraction of the computation cost of a full calculation on multilayer systems .
we show that using this scheme , one is able to capture the long range correlation effects within gw which leads to a significant renormalization of the gap even when dft finds the layers to be non - interacting at larger interlayer spacings .
this scheme can also be useful to study the variation in band structure as a function of stacking between the layers .
furthermore , it can be extended to study the nature of interaction between layers in heterostructures of these materials .
it paves a way to derive properties of a heterostructure from just unit cell calculations on the constituent materials .
this scheme is a promising tool to study multilayers and layer dependence of properties in the growing family of 2d layered materials .
45ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1073/pnas.0502848102 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.136805 [ * * , ( ) ] link:\doibase 10.1021/nl903868w [ * * , ( ) ] link:\doibase 10.1038/nnano.2012.193 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.111.106801 [ * * , ( ) ] link:\doibase 10.1103/physrevb.89.235319 [ * * , ( ) ] link:\doibase 10.1103/physrevb.83.245213 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.111.216805 [ * * , ( ) ] link:\doibase 10.1021/nn501226z [ * * , ( ) ] link:\doibase 10.1038/nnano.2010.279 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1063/1.3672219 [ * * , ( ) ] link:\doibase 10.1103/physrevb.93.075440 [ * * , ( ) ] link:\doibase 10.1103/physrevb.90.205420 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.86.241201 [ * * , ( ) ] link:\doibase 10.1021/acs.jpclett.5b02687 [ * * , ( ) ] link:\doibase 10.1021/acs.jpclett.6b00693 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.111.036104 [ * * , ( ) ] @noop * * , ( ) in @noop _ _ , , vol . , ( , ) pp
. link:\doibase 10.1103/physrevb.34.5390 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.55.1418 [ * * , ( ) ] link:\doibase 10.1021/acs.nanolett.5b00160 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevlett.110.146801 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1021/acs.nanolett.6b02084 [ * * , ( ) ] link:\doibase 10.1103/physrev.136.b864 [ * * , ( ) ] link:\doibase 10.1103/physrev.140.a1133 [ * * , ( ) ] link:\doibase 10.1557/jmr.2016.99 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.115.119901 [ * * , ( ) ] link:\doibase 10.1103/physrevb.93.235435 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.23.5048 [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.472933 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.cpc.2011.12.006 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.107.186404 [ * * , ( ) ] link:\doibase 10.1103/physrevb.73.233103 [ * * , ( ) ] link:\doibase 10.1103/physrevb.87.165124 [ * * , ( ) ] link:\doibase 10.1103/physrevb.89.075409 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.046802 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.97.216405 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.108.087402 [ * * , ( ) ] | we study the origin of layer dependence in band structures of two - dimensional materials .
we find that the layer dependence , at the density functional theory ( dft ) level , is a result of quantum confinement and the non - linearity of the exchange - correlation functional .
we use this to develop an efficient scheme for performing dft and gw calculations of multilayer systems .
we show that the dft and quasiparticle band structures of a multilayer system can be derived from a single calculation on a monolayer of the material .
we test this scheme on multilayers of mos@xmath0 , graphene and phosphorene .
this new scheme yields results in excellent agreement with the standard methods at a fraction of the computation cost .
this helps overcome the challenge of performing fully converged gw calculations on multilayers of 2d materials , particularly in the case of transition metal dichalcogenides which involve very stringent convergence parameters . |
the _ sparse representation _ problem involves solving the system of linear equations @xmath0 where @xmath1 is assumed to be @xmath2-sparse ; i.e. @xmath3 is allowed to have ( at most ) @xmath2 non - zero entries .
the matrix @xmath4 is typically referred to as the _ dictionary _ with @xmath5 elements or _
atoms_. it is well - known that @xmath3 can be uniquely identified if @xmath6 satisfies the so called _ _ spark condition _
_ columns of @xmath6 are linearly independent . ] .
meanwhile , there exist tractable and efficient convex relaxations of the combinatorial problem of finding the ( unique ) @xmath2-sparse solution of @xmath7 with provable recovery guarantees @xcite . a related problem is _ dictionary learning _ or _
sparse coding _ @xcite which can be expressed as a sparse factorization @xcite of the data matrix @xmath8 ( where both @xmath6 and @xmath9 are assumed unknown ) given that each column of @xmath10 is @xmath2-sparse and @xmath6 satisfies the spark condition as before .
a crucial question is how many data samples ( @xmath11 ) are needed to _ uniquely _ identify @xmath6 and @xmath10 from @xmath12 ? unfortunately , the existing lower bound is ( at best ) exponential @xmath13 assuming an equal number of data samples over each @xmath2-sparse support pattern in @xmath10 @xcite . in this paper
, we address a more challenging problem .
in particular , we are interested in the above sparse matrix factorization problem @xmath8 ( with both sparsity and spark conditions ) when only @xmath14 random linear measurements from each column of @xmath12 is available .
we would like to find lower bounds for @xmath11 for the ( partially observed ) matrix factorization to be unique .
this problem can also be seen as recovering both the dictionary @xmath6 and the sparse coefficients @xmath10 from compressive measurements of data .
for this reason , this problem has been termed _ blind compressed sensing _ ( bcs ) before @xcite , although the end - goal of bcs is the recovery of @xmath12 .
we start by establishing that the uniqueness of the learned dictionary over random data measurements is a sufficient condition for the success of bcs .
perfect recovery conditions for bcs are derived under two different scenarios . in the first scenario ,
fewer random linear measurements are available from each data sample .
it is stated that having access to a large number of data samples compensates for the inadequacy of sample - wise measurements . meanwhile , in the second scenario , it is assumed that slightly more random linear measurements are available over each data sample and the measurements are partly fixed and partly varying over the data .
this measurement scheme results in a significant reduction in the required number of data samples for perfect recovery .
finally , we address the computational aspects of bcs based on the recent non - iterative dictionary learning algorithms with provable convergence guarantees to the generating dictionary .
bcs was initially proposed in @xcite where it was assumed that , for a given random gaussian sampling matrix @xmath15 ( @xmath16 ) , @xmath17 is observed .
the conclusion was that , assuming the factorization @xmath8 is unique , @xmath18 factorization would also be unique with a high probability when @xmath6 is an orthonormal basis
. however , it would be impossible to recover @xmath6 from @xmath19 when @xmath14 .
it was suggested that structural constraints be imposed over the space of admissible dictionaries to make the inverse problem well - posed .
some of these structures were sparse bases under known dictionaries , finite set of bases and orthogonal block - diagonal bases @xcite .
while these results can be useful in many applications , some of which are mentioned in @xcite , they do not generalize to unconstrained overcomplete dictionaries .
subsequently , there has been a line of empirical work on showing that dictionary learning from compressive data a sufficient step for bcs can be successful given that a different sampling matrix is employed for each data sample is no longer valid which is possibly a reason for the lack of a theoretical extension of bcs to this case . ]
( i.e. each column of @xmath12 ) .
for example , @xcite uses a modified k - svd to train both the dictionary and the sparse coefficients from the incomplete data .
meanwhile , @xcite use generic gradient descent optimization approaches for dictionary learning when only random projections of data are available .
the empirical success of dictionary learning with partial as well as compressive or projected data triggers more theoretical interest in finding the uniqueness bounds of the unconstrained bcs problem .
finally , we must mention the theoretical results presented in the pre - print @xcite on bcs with overcomplete dictionaries while @xmath10 is assumed to lie in a structured union of disjoint subspaces @xcite .
it is also proposed that the results of this work extend to the generic sparse coding model if the ` one - block sparsity ' assumption is relaxed .
we argue that the main theoretical result in this pre - print is incomplete and technically flawed as briefly explained here . in the proof of theorem 1 of @xcite
, it is proposed that ( with adjustment of notation ) _ `` assignment [ of @xmath12 s columns to rank-@xmath20 disjoint subsets ] can be done by the ( admittedly impractical ) procedure of testing the rank of all possible @xmath21 matrices constructed by concatenating subsets of @xmath22 column vectors , as assumed in @xcite''_. however , it is ignored that the entries of @xmath12 are missing at random and the rank of an incomplete matrix can not be measured . as it becomes more clear later , the main challenge in the uniqueness analysis of unconstrained bcs is in addressing this particular issue .
two strategies to tackle this issue that are presented in this paper are : 1 ) increasing the number of data samples and 2 ) designing and employing measurement schemes that preserve the low - rank structure of @xmath12 s sub - matrices .
this paper is organized as follows . in section [ sec : problem ] , we provide the formal problem definition for bcs .
our main results are presented in section [ sec : main ] .
we present the proofs in section [ sec : proofs ] .
practical aspects of bcs are treated in section [ sec : algorithm ] where we explain how provable dictionary learning algorithms , such as @xcite , can be utilized for bcs . finally , we conclude the paper and present future directions in section [ sec : conclusion ] .
our general convention throughout this paper is to use capital letters for matrices and small letters for vectors and scalars . for a matrix @xmath9
, @xmath23 denotes its entry on row @xmath24 and column @xmath25 , @xmath26 denotes its @xmath24th column and @xmath27 denotes its column - major vectorized format .
the inner product between two matrices @xmath6 and @xmath28 ( of the same sizes ) is defined as @xmath29 .
let @xmath30 denote the smallest number of @xmath6 s columns that are linearly dependent .
@xmath6 is @xmath31-coherent if @xmath32 we have @xmath33 .
finally , let @xmath34\coloneqq\{1,2,\dots , m\}$ ] and let @xmath35\choose k}$ ] denote the set of all subsets of @xmath34 $ ] of size @xmath2 .
construct the data matrix @xmath36 by concatenating @xmath11 signal vectors @xmath37 ( for @xmath25 from 1 to @xmath11 ) . throughout this paper , we make the following assumptions about the sampling operator and the data sparsity .
it must be noted that the following assumptions over the sparse coding of @xmath12 are minimal among existing sparse coding assumptions for provable uniqueness guarantees ; see e.g. @xcite .
suppose @xmath38 linear measurements are taken from each signal @xmath37 as in @xmath39 where @xmath40 is referred to as the sampling matrix .
we could also represent the measurements as a linear projection of the signal onto the row - space of the sampling matrix can be computed from @xmath41 using the relationship @xmath42 .
therefore , @xmath41 and @xmath43 carry the same amount of information about @xmath44 given the sampling matrix @xmath45 .
] : @xmath46 we will use @xmath47^t\in\mathbb{r}^{pn}$ ] to denote the observations and @xmath48 to denote the projected matrix that is a concatenation of all @xmath41 . specifically ,
when entries of each @xmath45 are drawn independently from a random gaussian distribution with mean zero and variance @xmath49 , we use the notations @xmath50 and @xmath51 .
assume @xmath8 where @xmath4 denotes the dictionary ( @xmath52 in the overcomplete setting ) and @xmath9 is a sparse matrix with exactly @xmath2 non - zero entries per column and @xmath53 .
additionally , assume that each column of @xmath10 is randomly drawn by first selecting its support @xmath54\choose k}$ ] uniformly at random and then filling the support entries with random i.i.d .
values uniformly drawn from a bounded interval , e.g. @xmath55\subset\mathbb{r}$ ] .
we denote by @xmath56 the set of feasible @xmath12 under the described sparse coding model .
note that the assumption @xmath53 is necessary to ensure a unique @xmath10 even when @xmath6 is known and fixed . as noted and proved in @xcite , when @xmath57 , with probability one , no subset of @xmath2 ( or less ) columns of @xmath12 is linearly dependent . also with probability one ,
if a subset of @xmath58 columns of @xmath12 are linearly dependent , then all of the @xmath58 columns must have the same support .
given the above definitions , we can now formally express the problem definition for bcs : recover @xmath57 from @xmath59 given @xmath60 , @xmath61 and @xmath2 .
our results throughout this paper are mainly developed for the class of gaussian measurements @xmath62 .
however , it is not difficult to extend these results to the larger class of continuous sub - gaussian distributions for @xmath60 .
to start with , assume that there are exactly @xmath63 columns in @xmath10 for each support pattern @xmath64 where @xmath65\choose k}$ ] .
for better understanding and without loss of generality , one can assume that the data samples are ordered according to the following sketch for @xmath10 : @xmath66 \noalign{\vspace{-2\normalbaselineskip } } \multicolumn{7}{c}{n=\ell|\mathcal{s}|\mbox { samples}}\\ \multicolumn{7}{c}{$\downbracefill$}\\ \multicolumn{3}{c}{\mbox{group } \#1\;\;\ ; } & \multicolumn{1}{c } { } & \multicolumn{3}{c}{\mbox{group } \#|\mathcal{s}|\;\;\;}\\ \multicolumn{3}{c}{\mbox{support : } s^{1 } } & \multicolumn{1}{c } { } & \multicolumn{3}{c}{\mbox{support : } s^{|\mathcal{s}| } } \\ x_1 & \dots & x_\ell & \dots & x_{n-\ell + 1 } & \dots & x_n \\
\multicolumn{3}{c}{$\upbracefill$ } & \multicolumn{1}{c}{}&\multicolumn{3}{c}{$\upbracefill$ } \\
\multicolumn{3}{c}{\ell\mbox { samples } } & \multicolumn{1}{c } { } & \multicolumn{3}{c}{\ell\mbox { samples } } \\ \noalign{\vspace{-2\normalbaselineskip } } \end{pmatrix } \vspace{2\normalbaselineskip } \ ] ] the best known bound for @xmath63 , for the factorization @xmath8 to be unique ( with probability one ) under the specified random sparse coding model , is @xmath67 .
this results in an exponential sample complexity @xmath13 .
specifically , it is said that ` @xmath8 factorization is unique ' if there exist a diagonal matrix @xmath68 and a permutation matrix @xmath69 such that for any other feasible factorization @xmath70 , we have @xmath71 .
clearly , this ambiguity makes it more challenging to prove the uniqueness of the dictionary learning problem .
meanwhile , authors in @xcite propose a strategy for handling the permutation and scaling ambiguity which is reviewed in lemma [ lem : apd ] . through the following lemma
, we can establish that the uniqueness of the learned dictionary is a sufficient condition for the success of bcs ( proof is provided in appendix ) .
suppose for every pair @xmath72 that satisfy @xmath73 with @xmath74 , @xmath71 for some diagonal matrix @xmath75 and permutation matrix @xmath69 .
then @xmath76 with probability one .
[ lem : forgot ] briefly speaking , existing uniqueness results exploit the fact that the rank of each group of columns in the above sketch is bounded above by @xmath2 .
this makes it possible to uniquely identify groups of samples that share the same support pattern .
meanwhile , when only @xmath59 is available , it might not be possible to uniquely identify these groups .
nevertheless , it is noted in @xcite that @xmath77 ensures uniqueness without the need for grouping , at the cost of significantly increasing the required number of data samples ( compared to @xmath67 ) . in our initial bcs uniqueness result
, we use the pigeon - hole strategy of @xcite which results in a less practical bound @xmath78 even when @xmath12 is completely observed . ] . yet , it is interesting to explore the implications of a finite @xmath11 that ensures a successful bcs for the general sparse coding model .
the cs theory requires the complete knowledge of @xmath6 to uniquely recover @xmath10 and @xmath12 from @xmath59 .
meanwhile , our results assert that @xmath6 , @xmath10 and @xmath12 can be uniquely identified from @xmath59 given a large but finite number of samples @xmath11 .
necessary proofs for the results of this section are presented in the following section .
assume @xmath74 and there are exactly @xmath63 columns in @xmath10 for each @xmath64
. then @xmath57 can be perfectly recovered from @xmath50 with probability one given that @xmath79 .
[ lem : gell-2k ] with probability at least @xmath80 , @xmath57 can be perfectly recovered from @xmath50 given that @xmath74 and @xmath81 .
[ cor : gn-2k ] aside from the intellectual implications of theorem [ lem : gell-2k ] and corollary [ cor : gn-2k ] discussed above , the stated bounds for @xmath63 and @xmath11 are clearly not very practical . to reduce these bounds while guaranteeing the success of bcs
, we introduce a _ hybrid measurement _ scheme that we explain below . *
( hybrid gaussian measurement ) * in a hybrid measurement scheme , @xmath82 $ ] where @xmath83 stands for the fixed part of sampling matrix and @xmath84 stands for the varying part of the sampling matrix .
the total number of measurements per column is @xmath85 . in a hybrid gaussian measurement scheme , @xmath86 and @xmath87 through @xmath88
are assumed to be drawn independently from an i.i.d .
zero - mean gaussian distribution with variance @xmath49 .
the observations corresponding to @xmath86 and @xmath89 s are denoted by @xmath90 and @xmath91 respectively . as mentioned earlier ,
the hybrid measurement scheme was designed to reduce the required number of data samples for perfect bcs recovery . in particular , as formalized in lemma [ lem : rank - check ] , the fixed part of the measurements is designed to retain the low - rank structure of each @xmath2-dimensional subspace associated with a particular @xmath64 . meanwhile ,
the varying part of the measurements is essential for the uniqueness of the learned dictionary .
assume @xmath92 and there are exactly @xmath63 columns in @xmath10 for each @xmath64
. then @xmath57 can be perfectly recovered from hybrid gaussian measurements with probability one given that @xmath93 .
[ cor : gell-4k ] similar to the statement of corollary [ cor : gn-2k ] , it can be stated that bcs with hybrid gaussian measurement succeeds with probability at least @xmath80 given that @xmath94 .
the proof follows the proof of corollary [ cor : gn-2k ] .
although we mainly follow the stochastic approach of @xcite in this paper , we could also employ the deterministic approach of @xcite to arrive at the uniqueness bound in theorem [ cor : gell-4k ] . in @xcite , an algorithm ( which is not necessarily practical )
is proposed to uniquely recover @xmath6 and @xmath10 from @xmath12 .
this algorithm starts by finding subsets of size @xmath63 of @xmath12 s columns that are linearly dependent by testing the rank of every subset .
dismissing the degenerate possibilities are dismissed by adding extra assumptions in the deterministic sparse coding model .
meanwhile , as pointed out in @xcite , such degenerate instances of @xmath10 would have a probability measure of zero in a random sparse coding model ] , these detected subsets would correspond to samples with the same support pattern in @xmath10 . under the assumptions in theorem
[ cor : gell-4k ] , it is possible to test whether @xmath63 columns in @xmath12 are linearly dependent ( with probability one ) , as a consequence of lemma [ lem : rank - check ] in the following section . until now , our goal was to show that @xmath6 ( and subsequently @xmath10 ) is unique given only cs measurements .
as we mentioned before , uniqueness of @xmath6 is a _ sufficient _ condition for the success of bcs .
consider the scenario where not all support patterns @xmath64 are realized in @xmath10 or for some there is not enough samples to guarantee recovery .
for such scenarios , we present the following theorem .
assume @xmath92 and let @xmath95 where @xmath96 and @xmath97 denotes the set of indices of columns of @xmath10 with support @xmath98 .
then , under hybrid gaussian measurement , @xmath99 for all @xmath100 can be perfectly recovered with probability one .
[ the : latest ]
the following crucial lemma from @xcite handles the permutation ambiguity of sparse coding .
assume @xmath53 for @xmath4 and let @xmath65\choose k}$ ] . if there exists a mapping @xmath101 such that @xmath102 then there exist a permutation matrix @xmath69 and a diagonal matrix @xmath75 such that @xmath71 .
[ lem : apd ] the following lemma from random matrix theory , along with lemma [ lem : apd ] , are the main ingredients of our first main result ( proof is provided in the appendix ) .
assume @xmath103 are rank-@xmath2 matrices and @xmath104 is a gaussian measurement operator with @xmath105 .
if @xmath106 , then @xmath107 with probability one
. [ lem : manifold ] assume @xmath108 is an alternate factorization that satisfies @xmath109 and @xmath73 .
we will prove @xmath71 for some diagonal @xmath75 and some permutation matrix @xmath69 using lemma [ lem : apd ] .
consider a particular support pattern @xmath64 and let @xmath110 $ ] denote the set of indices of @xmath10 s columns that have the sparsity pattern @xmath98 . by definition , @xmath111 where @xmath112 .
due to the pigeon - hole principle , there must be at least @xmath113 columns within @xmath114 that share some particular support pattern @xmath115 .
in other words , if @xmath116 denotes the set of indices of @xmath117 s columns that have the support pattern @xmath118 , then @xmath119 . for simplicity , denote @xmath120 .
clearly , @xmath121 ( because @xmath122 ) , and we have @xmath123 according to lemma [ lem : manifold ] , if @xmath124 or equivalently @xmath125 , then @xmath126 with probability one .
meanwhile , since @xmath127 , @xmath126 necessitates that @xmath128 finally , since @xmath6 satisfies the spark condition , it is not difficult to see that @xmath129 is a bijective map . to explain more , assume there exists some @xmath130 such that @xmath131 combining with ( [ eq : map2 ] ) we arrive at @xmath132 which contradicts the spark condition for @xmath6 for @xmath130 .
therefore , @xmath133 must be injective . now
, since @xmath134 is a finite set and @xmath133 is an injective mapping from @xmath134 to itself , it must also be surjective and , thus , bijective . in order to have at least @xmath63 columns in @xmath10 for each support @xmath64 in the random sparse coding model @xmath56
, we must have more than just @xmath135 data samples . the following result from @xcite
quantifies the number of required data samples to ensure at least @xmath63 columns per each @xmath64 with a tunable probability of success . for a randomly generated @xmath10 with @xmath136 and @xmath137 $ ] , with probability at least @xmath80 , there are at least @xmath138 columns for each support pattern @xmath64 .
[ lem : beta ] proof is fairly trivial . according to lemma [ lem : beta ] ,
we need @xmath139 samples to guarantee that with probability at least @xmath80 there are at least @xmath63 samples in @xmath10 for each support pattern @xmath64 . in theorem [ lem :
gell-2k ] we established that @xmath79 guarantees the success of bcs under gaussian sampling .
therefore , @xmath140 guarantees the desired uniqueness . in order to prove the results for the hybrid measurement scheme
, we present the following lemma which is proved in the appendix .
assume @xmath83 is drawn from an i.i.d .
zero - mean gaussian distribution ( with @xmath141 ) .
let @xmath142 denote the columns of @xmath12 indexed by the set @xmath143 .
if @xmath144 , then @xmath145 with probability one
. [ lem : rank - check ] assume @xmath108 is an alternate factorization that satisfies @xmath109 , @xmath146 and @xmath147 .
also assume @xmath148 and @xmath149 .
consider a particular support pattern @xmath115 and let @xmath150 $ ] denote the set of indices of @xmath117 s columns that have the same sparsity pattern @xmath118 .
clearly , @xmath151 therefore , if @xmath152 , then @xmath153 and according to lemma [ lem : rank - check ] : @xmath154 with probability one .
hence , @xmath155 . again using lemma [ lem : rank - check ] with @xmath152 , @xmath156 with probability one . therefore
, all the columns in @xmath157 must have the same support , namely @xmath98 .
note that since @xmath158 , @xmath159 .
meanwhile , @xmath160 necessitates that @xmath161 for every @xmath115 .
therefore , @xmath162 .
now , given @xmath163 according to lemma [ lem : manifold ] , if @xmath164 , then @xmath126 with probability one . meanwhile , since @xmath165 , @xmath126 necessitates that @xmath166 finally , since @xmath6 satisfies the spark condition , @xmath129 is a bijective map and @xmath71 for some diagonal @xmath75 and permutation matrix @xmath69 according to lemma [ lem : apd ] .
recall that for every @xmath100 we have @xmath167 .
assume @xmath148 and @xmath149 as before .
having @xmath168 allows testing whether a subset of @xmath58 columns of @xmath12 are linearly dependent ( have a rank of @xmath2 ) with probability one .
therefore , by doing an exhaustive search among every sub - matrix @xmath169 with @xmath170\choose k+1}$ ] , we are able to find subsets of @xmath97 ( of size @xmath58 ) if @xmath171 .
moreover , we can combine and complete these subsets to uniquely identify every rank-@xmath2 sub - matrix @xmath99 with @xmath171 .
now , among these sub - matrices , those with @xmath172 can be recovered perfectly ( with probability one ) since , for any rank-@xmath2 matrices @xmath99 and @xmath173 , @xmath174 with @xmath175 or @xmath176 implies @xmath177 according to lemma [ lem : manifold ] .
recall that in the dictionary learning ( dl ) problem , the data matrix @xmath178 is given where @xmath179 and the task is to factorize @xmath180 such that @xmath181 for some permutation matrix @xmath69 and diagonal matrix @xmath75 .
unfortunately , the corresponding optimization problem is non - convex ( even with @xmath182 relaxation ) .
the majority of existing dl algorithms are based on the iterative scheme of starting from an initial state @xmath183 and alternating between updating @xmath184 while keeping @xmath185 fixed and updating @xmath186 while keeping @xmath184 fixed , each corresponding to a convex problem .
it has been recently shown that if the initial dictionary @xmath187 is sufficiently close @xcite . ] to @xmath188 for some @xmath69 and @xmath75 , then the iterative algorithm converges to @xmath188 under certain incoherency assumptions about @xmath189 @xcite .
similar guarantees have been derived for the well - known k - svd algorithm @xcite .
furthermore , dl from incomplete or corrupt data has also been tackled in several studies . in particular , dl from compressive measurements has been addressed in @xcite where different iterative dl algorithms are modified to accommodate the compressive measurements . in some cases ,
these modifications have been justified by showing that the output of each iteration does not significantly deviate from the reference output based on the complete data . however , to best of our knowledge , there are no convergence guarantees to @xmath188 for these iterative algorithms . as we mentioned before , a successful dl from compressive measurements is a sufficient condition for a successful bcs . in this section ,
we plan to investigate the utility of a recently proposed ( non - iterative ) dl algorithm @xcite with guarantees for the approximate recovery of @xmath188 for an incoherent @xmath189 .
one would hope that @xmath188 can be approximated from @xmath12 with fewer data samples than is required for the exact identification of @xmath188 which was the topic of previous sections .
below , we review the main result of @xcite and analyze the performance of their dl algorithm if only hybrid gaussian measurements were available .
recall that in our bcs measurement scheme , @xmath190 fixed and @xmath191 varying linear measurements are taken from each sample for a total of @xmath192 linear measurements ( per sample ) . before presenting their result , we need to introduce some new notation as well as modifications to the sparse coding model to reflect the model used in @xcite .
in particular , let @xmath193 denote the random vector of sparse coefficients where its distribution class @xmath194 is defined below .
hence , each @xmath195 denotes an outcome of @xmath196 .
also , let @xmath197 denote the random variable associated with the @xmath24th entry of @xmath196 .
( distribution class @xmath194 ) the distribution is in class @xmath194 if @xmath24 ) @xmath198\cup [ 1,c]$ ] and @xmath199=0 $ ] .
@xmath200 ) conditioned on any subset of coordinates in @xmath196 being non - zero , the values of @xmath197 are independent of each other .
distribution has _ bounded @xmath63-wise moments _ if the probability that @xmath196 is non - zero in any subset @xmath98 of @xmath63 coordinates is at most @xmath201 times @xmath202 $ ] where @xmath203 .
similar to @xcite , in the rest of paper we will assume @xmath204 .
derived results generalize to the case @xmath205 by loosing constant factors in guarantees .
two dictionaries @xmath206 are column - wise @xmath207-close , if there exists a permutation @xmath133 and @xmath208 such that @xmath209\colon \|a_i-\theta_i b_{\pi(i)}\|_2\leq \epsilon$ ] .
when talking about two dictionaries @xmath6 and @xmath28 that are @xmath207-close , we always assume the columns are ordered and scaled correctly so that @xmath210 .
there is a polynomial time algorithm to learn a @xmath31-coherent dictionary @xmath6 from random samples . with high probability ,
the algorithm returns a dictionary @xmath211 that is column - wise @xmath207-close to @xmath6 given random samples of the form @xmath212 , where @xmath196 is drawn from a distribution in class @xmath194 .
specifically , if @xmath213 and the distribution has bounded @xmath63-wise moments , @xmath214 is a constant only depending on @xmath63 , then the algorithm requires @xmath215 samples and runs in time @xmath216 . [
the : arora ] this algorithm , which has fundamental similarities with a concurrent work @xcite , consists of two main stages : @xmath24 ) _ data clustering _ :
the connection graph is built where each node corresponds to a column of @xmath12 and an edge between @xmath217 and @xmath44 implies their supports @xmath218 and @xmath219 have a non - empty intersection .
then , an overlapping clustering procedure is performed over the connection graph to find overlapping maximal cliques ( with missing edges ) .
@xmath200 ) _ dictionary recovery _ : every cluster in the connection graph represents the set of samples associated with a single dictionary atom . after finding these clusters in the connection graph ,
each atom is approximated by the principal eigenvector of the covariance matrix for the data samples in its corresponding cluster .
there are two challenges in extending the above result to the bcs framework : @xmath24 ) during generation of the connection graph from data and @xmath200 ) during computation of the principal eigenvector of the data covariance matrix .
we address these challenges separately in the following subsections . for building the connection graph , we use the fixed part of the hybrid measurements , i.e. @xmath220 with @xmath83 drawn from a gaussian distribution .
computation of the connection graph in @xcite relies on the following lemma .
suppose @xmath221 for large enough @xmath222 ( depending on @xmath223 in the definition of @xmath194 ) .
then , if @xmath218 and @xmath219 are disjoint , with high probability @xmath224 . without going into the details of the clustering algorithm of @xcite , we study the conditions under which the connection graph does not change when only @xmath190 linear measurements from each data sample is given .
let @xmath225 be @xmath226-coherent .
it is not hard to see from the above lemma that if @xmath227 , then with high probability for disjoint @xmath218 and @xmath219 , @xmath228 .
to establish a relationship between @xmath226 , @xmath31 and @xmath190 , we use the following result from @xcite .
let @xmath229 with @xmath230 .
assume @xmath231 is a random matrix with independent @xmath232 entries .
then , for all @xmath233 @xmath234\leq 2\exp ( -n\frac{t^2}{c_1+c_2t})\ ] ] with @xmath235 and @xmath236 .
[ lem : holger ] assume @xmath83 has i.i.d entries from @xmath237 .
let @xmath6 be @xmath31-coherent and @xmath238 be @xmath226-coherent .
then , @xmath239\leq 2\exp ( -p_f\frac{t^2}{c_1+c_2t})\ ] ] with @xmath240 and @xmath241 specified in lemma [ lem : holger ] .
[ cor : mu ] note that the variance of @xmath86 s entries does not have an effect on @xmath226 due to the normalization in the definition of the coherency and we could assume @xmath86 s entries have variance @xmath49 as before .
we exploit lemma [ lem : holger ] by replacing @xmath242 and @xmath243 and @xmath244 .
proof is complete by noticing that @xmath245\leq\mathbb{p}[|\mu_f-\mu|\geq t]$ ] based on corollary [ cor : mu ] , it can be deduced that with high probability @xmath246 .
therefore , replacing @xmath247 in the original theorem [ the : arora ] with @xmath248 introduces slightly stronger sparsity requirement for the success of the algorithm . at this stage , we only exploit the varying part of the measurements @xmath249 and use @xmath250 in place of @xmath191 for simplicity .
let @xmath251 be the @xmath61 discovered overlapping clusters from the previous stage and define the empirical covariance matrix @xmath252 for the cluster @xmath24 .
the svd approach ) selective averaging and @xmath200 ) the svd - based approach .
we selected to work with the svd approach due to its more abstract and versatile nature .
] of @xcite estimates @xmath253 by @xmath254 which is the principal eigenvector of @xmath255 .
let @xmath256 denote the empirical covariance matrix resulting from the compressive measurements where @xmath257 as before .
similarly , let @xmath258 denote the principal eigenvector of @xmath259 .
our goal in this section is to show that @xmath260 is bounded by a small constant for finite @xmath11 and approaches zero for large @xmath11 . for this purpose
, we use the recent results from the area of _ subspace learning _ , specifically , subspace learning from compressive measurements @xcite . a critical factor in estimation accuracy of the principle eigenvector of a _ perturbed _ covariance matrix is the _ eigengap _ between the principal and the second eigenvalues of the original covariance matrix .
this is a well - known result from the works of chandler davis and william kahan known as the davis - kahan sine theorem @xcite .
consider the following notation .
let @xmath261 and @xmath262 denote projection operators onto the principal @xmath2-dimensional subspaces of @xmath263 and @xmath264 respectively ( i.e. the projection onto the top-@xmath2 eigenvectors ) .
let @xmath265 denote the spectral norm of the difference between @xmath261 and @xmath262 .
define the eigengap @xmath266 as the distance between the @xmath2th and @xmath58st largest eigenvalues of @xmath263 .
suppose @xmath263 is computed from at least @xmath63 data samples ( @xmath267 for all @xmath24 ) .
moreover , assume the data samples have bounded @xmath268 norms , i.e. @xmath269\colon\|y_j\|_2 ^ 2\leq \eta$ ] for some positive @xmath270 . with probability
at least @xmath271 @xmath272 so that one can achieve @xmath273 provided that @xmath274 [ lem : akshay ] below , we present a customization of lemma [ lem : akshay ] for the @xmath268 error of the principal eigenvector estimator .
let @xmath254 and @xmath258 represent the principal eigenvectors of @xmath255 and @xmath259 respectively .
with probability at least @xmath271 for all @xmath275 $ ] @xmath276 [ lem : temp ] clearly , @xmath277 and @xmath278 . as we mentioned in the definition of @xmath207-closeness
, @xmath279 is implicit in the error expression @xmath280 requiring that @xmath281 and consequently @xmath282 .
also note that , by definition , for any @xmath283 @xmath284 now let @xmath285 . then @xmath286 therefore @xmath287 and the rest follows from lemma [ lem : akshay ] . to obtain
a lower - bound for the eigengap @xmath288 we need to review some of the intermediate results in @xcite .
in fact , we compute a lower - bound for @xmath289 of @xmath290 which serves as a close approximation of @xmath288 when the number of data samples @xmath63 is large . for every @xmath275 $ ] , let @xmath291 be the distribution conditioned on @xmath292 .
let @xmath293 for any unit - norm @xmath294 and let @xmath295 = 1+\sum_{j\neq i } \langle a_i , a_j \rangle ^2 \mathbb{e}_{\gamma_i}[\mathcal{x}_j^2]\ ] ] denote the _ projected variance _ of @xmath291 on the direction @xmath296 .
it is shown @xcite that generally @xmath297 \leq \alpha^2 r_i^2 + 2\alpha\sqrt{1-\alpha^2}\zeta + ( 1-\alpha^2)\zeta^2\ ] ] where @xmath298 .
the principal eigenvector of @xmath299 can be computed by finding the unit - norm @xmath294 that maximizes @xmath300 $ ] .
meanwhile , it has been established that for @xmath296 , @xmath300 = r_i^2 $ ] .
therefore , the range of @xmath301 for the principal eigenvector must satisfy the inequality ( for @xmath302 ) @xmath303 it is not difficult to show this range is @xmath304\ ] ] now , for the second eigenvector and eigenvalue pair we must find a unit - norm @xmath305 that satisfies @xmath306 and maximizes @xmath307 $ ] .
define @xmath308 .
it can be shown that @xmath309\ ] ] note that the first and the second largest eigenvalues correspond to projected variances of @xmath291 on the directions of @xmath294 and @xmath305 , respectively .
therefore , based on the derived ranges for @xmath301 and @xmath310 , we are able to find the following lower - bound for @xmath289 : @xmath311 note that @xmath312 becomes very small as the problem size ( @xmath313 , @xmath61 , @xmath11 ) becomes large , resulting in @xmath314 . therefore , given a sufficient number of samples , it can be guaranteed that @xmath258 is an accurate estimation of @xmath254 and , in turn , an accurate estimation of @xmath253 even when only @xmath14 measurements per sample is available . once the dictionary has been approximated to within a close distance from the optimal dictionary @xmath315 , iterative algorithms such as @xcite can assure convergence to a local optimum and therefore perfect recovery as suggested in @xcite .
finally , perfect recovery of the dictionary results in perfect recovery of @xmath10 and @xmath12 given the cs bounds for the number of measurements @xcite which are generally weaker than the stated bounds for the recovery of the dictionary .
in this work , we studied the conditions for perfect recovery of both the dictionary and the sparse coefficients from linear measurements of the data .
the first part of this work brings together some of the recent theories about the uniqueness of dictionary learning and the blind compressed sensing problem .
moreover , we described a ` hybrid ' random measurement scheme that reduces the theoretical bounds for the minimum number of data samples to guarantee a unique dictionary and thus perfect recovery for blind compressed sensing . in the second part , we discussed the algorithmic aspects of dictionary learning under random linear measurements .
it was shown that a polynomial - time algorithm can assure convergence to the generative dictionary given a sufficient number of data samples with high probability .
it would be interesting to explore dictionary learning and blind compressed sensing for non - gaussian random measurements .
in particular , when the data matrix is partially observed ( i.e. an incomplete matrix ) , data recovery becomes a matrix completion problem where the elements of the data matrix are assumed to lie in a union of interconnected rank-@xmath2 subspaces .
this is a subject of future work .
let @xmath316 . note that @xmath317 .
thus , @xmath318 . our goal is to show @xmath319 and thus @xmath320 . to prove @xmath319 , we must show that for every @xmath321 $ ] , @xmath322 results in @xmath323 with probability one . for simplicity , we omit the sample index @xmath25 in the rest of the proof .
let @xmath98 and @xmath324 respectively denote the sets of non - zero indices of @xmath3 and @xmath325 where @xmath326 . rewrite @xmath327 as @xmath328 .
note that @xmath329 is supported on @xmath330 where @xmath331 .
therefore , we must show that , with probability one , @xmath332\choose 2k}\colon rank(\phi a_t)=|t|\ ] ] necessitating @xmath333 or @xmath334 . since @xmath53 ,
every @xmath335 columns of @xmath6 are linearly independent and we are able to perform a gram - schmidt orthogonalization on @xmath336 to get @xmath337 where @xmath338 is orthonormal ( @xmath339 ) and @xmath340 is a full - rank square matrix .
hence , @xmath341 is distributed according to i.i.d .
gaussian and is full - rank with probability one @xcite .
we conclude the proof by noticing that @xmath342 since @xmath340 is a full - rank square matrix
. then @xmath348 .
specifically , under the gaussian measurement scheme for bcs , we have : @xmath349\in\mathbb{r}^{p n\times d n}\ ] ] where non - zero entries of @xmath350 are i.i.d .
gaussian with mean zero and variance @xmath49 .
the following result from @xcite gives the required number of linear measurements to guarantee ( with probability one ) that a rank-@xmath351 matrix does not fall into the null - space of the measurement operator .
let @xmath352 be a @xmath353-dimensional continuously differentiable manifold over the set of @xmath354 real matrices .
suppose we take @xmath355 linear measurements from @xmath356 .
assume there exists a constant @xmath357 such that @xmath358 for every @xmath12 with @xmath359 .
further assume that for each @xmath360 that the random variables @xmath361 are independent .
then with probability one , @xmath362 .
[ lem : r11 ] a careful inspection of the derivation of the above theorem in @xcite reveals that this result can be extended to include the manifolds over the set of rectangular matrices @xmath178 .
specifically , for the manifold over rank-@xmath351 @xmath363 matrices we have ( see @xcite for example ) @xmath364 .
let @xmath352 denote the manifold over the set of rank-@xmath351 @xmath363 matrices and let @xmath367 denote the manifold over the set of rank-@xmath368 @xmath363 matrices . also let @xmath343 with @xmath369 . then , for any @xmath370 , @xmath366 implies @xmath107 with probability one .
clearly , @xmath371 implies @xmath372 for any @xmath373 over the set of rank-@xmath374 @xmath363 matrices with @xmath375 . also note that @xmath376 , thus @xmath377 .
now , since @xmath378 and @xmath379 ( with probability one , according to lemma [ lem : r11 ] ) , we must have @xmath380 or @xmath107 with probability one .
it only remains to show that @xmath104 satisfies the requirements of lemma [ lem : r11 ] .
as noted in @xcite , the requirement @xmath381 requires that the densities of @xmath382 do not spike at the origin ; a sufficient condition for this to hold for every @xmath12 with @xmath359 is that each @xmath383 has i.i.d .
entries with a continuous density .
note that non - zero entries of @xmath384 are i.i.d .
gaussian and cover every column in @xmath12 . therefore
, none of the entries of @xmath385 would spike at the origin or equivalently there exists @xmath386 so that @xmath387 with @xmath388 given that the vector @xmath389 is drawn from a continuous distribution .
let @xmath390 and @xmath391 .
perform a gram - schmidt orthogonalization on @xmath169 to obtain @xmath392 where @xmath393 has orthogonal columns and @xmath394 is full - rank ; hence , given @xmath395 , we have @xmath396 .
note that , since @xmath397 is orthogonal and @xmath86 is i.i.d .
gaussian , @xmath398 is also i.i.d .
gaussian . hence , with probability one , @xmath398 is full - rank @xcite and @xmath399 . to conclude the proof ,
note that when @xmath400 , necessarily we have @xmath401 .
e. candes , j. romberg and t. tao , `` robust uncertainty principles : exact signal reconstruction from highly incomplete frequency information , '' _ ieee transactions on information theory _ , vol .
2 , pp . 489-509 , 2006 . | blind compressed sensing ( bcs ) is an extension of compressed sensing ( cs ) where the optimal sparsifying dictionary is assumed to be unknown and subject to estimation ( in addition to the cs sparse coefficients ) . since the emergence of bcs , dictionary learning , a.k.a .
sparse coding , has been studied as a matrix factorization problem where its sample complexity , uniqueness and identifiability have been addressed thoroughly . however , in spite of the strong connections between bcs and sparse coding , recent results from the sparse coding problem area have not been exploited within the context of bcs . in particular ,
prior bcs efforts have focused on learning constrained and complete dictionaries that limit the scope and utility of these efforts . in this paper
, we develop new theoretical bounds for perfect recovery for the general _ unconstrained _ bcs problem .
these unconstrained bcs bounds cover the case of overcomplete dictionaries , and hence , they go well beyond the existing bcs theory .
our perfect recovery results integrate the combinatorial theories of sparse coding with some of the recent results from low - rank matrix recovery .
in particular , we propose an efficient cs measurement scheme that results in practical recovery bounds for bcs .
moreover , we discuss the performance of bcs under polynomial - time sparse coding algorithms . |
the problem of community detection in networks has received wide attention @xcite .
it has proved to be a problem of remarkable subtlety , computationally challenging and with deep connections to other areas of research including machine learning , signal processing , and spin - glass theory .
a large number of algorithmic approaches to the problem have been considered , but interest in recent years has focused particularly on statistical inference methods @xcite , partly because they give excellent results , but also because they are mathematically principled and , at least in some cases , provably optimal @xcite . in this paper
we study two of the most fundamental community inference methods , based on the so - called stochastic block model or its degree - corrected variant @xcite .
we show that it is possible to map both methods onto the well - known minimum - cut graph partitioning problem , which allows us to adapt any of the large number of available methods for graph partitioning to solve the community detection problem . as an example
, we apply the laplacian spectral partitioning method of fiedler @xcite to derive a community detection method competitive with the best currently available algorithms in terms of both speed and quality of results .
the first method we consider is based on the stochastic block model , sometimes also called the planted partition model , a well studied model of community structure in networks @xcite .
this model supposes a network of @xmath0 vertices divided into some number of groups or communities , with different probabilities for connections within and between groups .
we will here focus on the simplest case of just two groups ( of any size , not necessarily equal ) . in the commonest version of the model edges
are placed independently at random between vertex pairs with probability @xmath1 for pairs in the same group and @xmath2 for pairs in different groups . in this paper
we use the slightly different poisson version of the model described in @xcite , in which we place between each pair of vertices a poisson - distributed number of edges with mean @xmath3 for pairs in the same group and @xmath4 for pairs in different groups . in essentially all
real - world networks the fraction of possible edges that are actually present in the network is extremely small ( usually modeled as vanishing in the large-@xmath0 limit ) , in which case the two versions of the model become indistinguishable , but the poisson version is preferred because its analysis is more straightforward . at its heart ,
the statistical inference of community structure is a matter of answering the following question : if we assume an observed network is generated according to our model , what then must the parameters of that model have been ?
in other words , what were the values of @xmath3 and @xmath4 used to generate the network and , more importantly , which vertices fell in which groups ?
even though the model is probably not a good representation of the process by which most real - world networks are generated , the answer to this question often gives a surprisingly good estimate of the true community structure . to answer the question , we make use of a maximum likelihood method .
let us label the two groups or communities in our model group 1 and group 2 , and denote by @xmath5 the group to which vertex @xmath6 belongs .
the edges in the network will be represented by an adjacency matrix having elements @xmath7 then the likelihood of generating a particular network or graph @xmath8 , given the complete set of group memberships , which we ll denote by the shorthand @xmath9 , and the poisson parameters , which we ll denote by @xmath10 , is @xmath11 where @xmath12 denotes the expected number of edges between vertices @xmath6 and @xmath13either @xmath3 or @xmath4 , depending on whether the vertices are in the same or different groups .
we are assuming there are no self - edges in the network
edges that connect vertices to themselves
so @xmath14 for all @xmath6 .
given the likelihood , one can maximize it to find the most likely values of the group labels and parameters , which can be done in a number of different ways . in ref .
@xcite , for example , the likelihood was maximized first with respect to the parameters @xmath3 and @xmath4 by differentiation . applying this method to eq .
gives most likely values of @xmath15 where @xmath16 and @xmath17 are the observed numbers of edges within and between groups respectively for a given candidate division of the network , and @xmath18 and @xmath19 are the numbers of vertices in each group . substituting these values back into eq .
gives the profile likelihood , which depends on the group labels only .
in fact , one typically quotes not the profile likelihood itself but its logarithm , which is easier to work with . neglecting an unimportant additive constant ,
the log of the profile likelihood for the present model is @xmath20 the communities can now be identified by maximizing this quantity over all possible assignments of the vertices to the groups .
this is still a hard task , however .
there are an exponentially large number of possible assignments , so an exhaustive search through all of them is unfeasible for all but the smallest of networks
. one can apply standard heuristics like simulated annealing to the problem , but in this paper we take a different approach . in the calculation above
, the likelihood is maximized over @xmath10 first , for fixed group assignments , then over the group assignments .
but we can also take the reverse approach , maximizing first over the group assignments , for given @xmath10 , and then over @xmath10 at the end .
this approach is attractive for two reasons .
first , as we will show , the problem of maximizing with respect to the group assignment when @xmath10 is given is equivalent to the standard problem of minimum - cut graph partitioning , a problem for which many excellent heuristics are already available .
second , after maximizing with respect to the group assignments the remaining problem of maximizing with respect to @xmath10 is a one - parameter optimization that can be solved trivially .
the net result is that the problem of maximum - likelihood community detection is reduced to one of performing a well - understood task graph partitioning plus one undemanding extra step .
the resulting algorithm is fast and , as we will see , gives good results .
so consider the problem of maximizing the likelihood , eq .
, with respect to the group labels @xmath5 , for given values of the parameters @xmath3 and @xmath4 . we will actually maximize the logarithm @xmath21 of the likelihood , @xmath22 , \label{eq : logl1}\ ] ] which gives the same result but is usually easier .
to proceed we write @xmath12 and @xmath23 as @xmath24 where @xmath25 is the kronecker delta . substituting these into eq . and
dropping overall additive and multiplicative constants , which have no effect on the position of the maximum , the log - likelihood can be rearranged to read @xmath26 where @xmath27 which is positive whenever @xmath28 , meaning we have traditional community structure in our network .
( it is possible to repeat the calculations for the case @xmath29 and derive methods for detecting such structure as well , although we will not do that here . )
the quantity @xmath30 is the cut size of the network partition represented by our two communities , i.e. , the number of edges connecting vertices in different communities , which we previously denoted @xmath17 , and @xmath31 where as previously @xmath18 and @xmath19 are the numbers of vertices in communities 1 and 2 .
thus we can also write the log - likelihood in the form @xmath32 the maximization of this log - likelihood corresponds to the minimization of the cut size , with an additional penalty term @xmath33 that favors groups of equal size .
this is similar , though not identical , to the so - called ratio cut problem , in which one minimizes the ratio @xmath34 , which also favors groups of equal size , although the nature of the penalty for unbalanced groups is different .
the catch with maximizing eq .
is that we do nt know the value of @xmath35 , which depends on the unknown quantities @xmath3 and @xmath4 via eq . , but we can get around this problem by the following trick . we first perform a limited maximization of in which the sizes @xmath18 and @xmath19 of the groups are held fixed at some values that we choose .
this means that the term @xmath36 is a constant and hence drops out of the problem and we are left maximizing only @xmath37 , or equivalently minimizing the cut - size @xmath17 .
this problem is now precisely the standard minimum - cut problem of graph partitioning the minimization of the cut size for divisions of a graph into groups of given sizes .
there are @xmath38 possible choices of the sizes of the two groups , ranging from putting all vertices in group 1 to all vertices in group 2 , and everything in between . if we solve the minimum - cut problem for each of these @xmath38 choices we get a set of @xmath38 solutions and we know that one of these must be the solution to our overall maximum likelihood problem .
it remains only to work out which one . but
choosing between them is easy , since we know that the true maximum also maximizes the profile likelihood , eq . .
so we can simply calculate the profile likelihood for each solution in turn and find the one that gives the largest result . in effect
, our approach narrows the exponentially large pool of candidate divisions of the network to a one - parameter family of just @xmath38 solutions ( parametrized by group size ) , from which it is straightforward to pick the overall winner by exhaustive search .
moreover , the individual candidate solutions are all themselves solutions of the standard minimum - cut partitioning problem , a problem that has been well studied for many years and about which a great deal is known @xcite . although partitioning problems are , in general , hard to solve exactly , there exist many heuristics that give good answers in practical situations .
the approach developed here allows us to apply any of these heuristics directly to the maximum - likelihood community detection problem . as an example of this approach , we demonstrate a fast and simple spectral algorithm based on the laplacian spectral bisection method for graph partitioning introduced by fiedler @xcite
a description of this method can be found , for example , in @xcite , where it is shown that a good approximation to the minimum - cut division of a network into two parts of specified sizes can be found by calculating the fiedler vector , which is the eigenvector of the graph laplacian matrix @xmath39 corresponding to the second smallest eigenvalue .
( the graph laplacian is the @xmath40 symmetric matrix @xmath41 , where @xmath42 is the adjacency matrix and @xmath43 is the @xmath40 diagonal matrix with @xmath44 equal to the degree of vertex @xmath6 . )
having calculated the fiedler vector one divides the network into groups of the required sizes @xmath18 and @xmath19 by inspecting the vector elements and assigning the @xmath18 vertices with the largest ( most positive ) elements to group 1 and the rest to group 2 .
although the method gives only an approximation to the global minimum - cut division , practical experience ( and some rigorous results ) show that it gives good answers under commonly occurring conditions @xcite .
a nice feature of this approach is that , in a single calculation , it gives us the entire one - parameter family of minimum - cut divisions of the network .
we need calculate the fiedler vector only once , sort its elements in decreasing order , then cut them into two groups in each of the @xmath38 possible ways and calculate the profile likelihood for the resulting divisions of the network .
the one with the highest score is ( an approximation to ) the maximum - likelihood community division of the network .
these developments are for the standard stochastic block model . as shown in ref .
@xcite , however , the standard block model gives poor results when applied to most real - world networks because the model fails to take into account the broad degree distribution such networks possess .
this problem can be fixed by a relatively simple modification of the model in which the expected number @xmath12 of edges between vertices @xmath6 and @xmath13 is replaced by @xmath45 where @xmath46 is the degree of vertex @xmath6 and @xmath12 again depends only on which groups the vertices @xmath6 and @xmath13 belong to .
all the developments for the standard block model above generalize in straightforward fashion to this `` degree - corrected '' model .
the log - likelihood and log - profile likelihood become @xmath47 where @xmath48 and @xmath49 are the sums of the degrees of the vertices in the two groups . in other words ,
the expressions are identical to those for the uncorrected model except for the replacement of the group sizes @xmath50 by @xmath51 .
the maximization of @xmath21 is thus once again reduced to a generalized minimum - cut partitioning problem , with a penalty term proportional to @xmath52 , which again favors balanced groups .
although we do nt know the value of @xmath35 , we can reduce the problem to a variant of the minimum - cut problem by the equivalent of our previous approach , holding @xmath48 and @xmath49 constant . and
again we can derive a spectral algorithm for this problem based on the graph laplacian . by a derivation analogous to that for the standard spectral method
we can show that a good approximation to the problem of minimum - cut partitioning with fixed @xmath51 ( as opposed to fixed @xmath50 ) is given not by the second eigenvector of @xmath39 but by the second eigenvector of the generalized eigensystem @xmath53 , where , as previously , @xmath43 is the diagonal matrix of vertex degrees .
once again we calculate the vector and split the vertices into two groups according to the sizes of their corresponding vector elements and once again this gives us a one - parameter family of @xmath38 candidate solutions from which we can choose an overall winner by finding the one with the highest profile likelihood , eq . .
vertices , generated using the standard ( uncorrected ) stochastic block model with equal group sizes of 5000 vertices each and a range of strengths of the community structure . defining @xmath54 , @xmath55 ,
the curves are ( top to bottom ) @xmath56 , 75 , 70 , 65 , and 60 , and @xmath57 .
the dashed vertical line indicates the true size of the planted communities .
the curves have been displaced from one another vertically for clarity .
the vertical axis units are arbitrary because additive and multiplicative constants have been neglected in the definition of the log - likelihood .
( b ) profile likelihoods for the same parameter values but unequal groups of size @xmath58 and @xmath59 .
( c ) the average fraction of vertices classified correctly for networks of @xmath60 vertices each and two equally sized groups .
each point is an average over 100 networks .
statistical errors are smaller than the points in all cases .
the vertical dashed line indicates the position of the `` detectability threshold '' at which community structure becomes formally undetectable @xcite . ]
we have tested this method on a variety of networks , and in practice it appears to work well . figure [ fig : synthetic ] shows results from tests on a large group of synthetic ( i.e. , computer - generated ) networks .
these networks were themselves generated using the standard stochastic block model ( which is commonly used as a benchmark for community detection @xcite ) .
the two left panels in the figure show the value of the profile likelihood for the families of @xmath38 candidate solutions generated by the spectral calculation for networks with two equally sized groups ( top ) and with unequal groups ( bottom ) . in each case
there is a clear peak in the profile likelihood at the correct group sizes , suggesting that the algorithm has correctly identified the group membership of most vertices .
the third panel in fig .
[ fig : synthetic ] tests this conclusion by calculating the fraction of correctly identified vertices as a function of the strength of the community structure for equally sized groups ( which is the most difficult case ) .
as the figure shows , the algorithm correctly identifies most vertices over a large portion of the parameter space .
the vertical dashed line represents the `` detectability threshold '' identified by previous authors @xcite , below which it is believed that every method of community detection must fail .
our algorithm fails below this point also , but appears to work well essentially all the way down to the transition , and there are reasons to believe this result to be exact , at least for networks that are not too sparse @xcite .
+ + figure [ fig : results ] shows the results of applications of the algorithm to two well - studied real - world networks , zachary s `` karate club '' network @xcite and adamic and glance s network of political blogs @xcite .
both are known to have pronounced community structure and the divisions found by our spectral algorithm mirror closely the accepted communities in both cases .
in addition to being effective , the algorithm is also fast .
the computation of the eigenvector can be done using , for instance , the lanczos method , an iterative method which takes time @xmath61 per iteration , where @xmath62 is the number of edges in the network .
the number of iterations required is typically small , although the exact number is not known in general .
the search for the division that maximizes the profile likelihood can also be achieved in @xmath61 time . of the @xmath38 different divisions of the network that must be considered ,
each one differs from the previous one by the movement of just a single vertex from one group to the other .
the movement of vertex @xmath6 between groups causes the quantities appearing in eq . to change according to @xmath63 where @xmath64 equals the number of edges between @xmath6 and vertices in group 1 minus the number between @xmath6 and vertices in group 2 .
these quantities and the resulting change in the profile likelihood can be calculated in time proportional to the degree of the vertex and hence all @xmath0 vertices can be moved in time proportional to the sum of all degrees in the network , which is equal to @xmath65 .
thus , to leading order , the total running time of the algorithm goes as @xmath62 times the number of lanczos iterations , the latter typically being small , and in practice the method is about as fast as the best competing algorithms .
in this paper we have shown that the widely - studied maximum likelihood method for community detection in networks can be reduced to a search through a small family of candidate solutions , each of which is itself the solution to a minimum - cut graph partitioning problem , which is a well studied problem about which much is known .
this mapping allows us to use trusted partitioning heuristics to solve the community detection problem . as an example
we have adapted the method of laplacian spectral partitioning to derive a spectral likelihood maximization algorithm and tested its performance on both synthetic and real - world networks . in terms of both accuracy and speed
we find the algorithm to be competitive with the best current methods .
a number of extensions of our approach would be possible , including extensions with more general forms for the parameters @xmath10 , such as different values of @xmath3 and @xmath4 for different groups , or extensions to more than two groups , but we leave these for future work .
the author would like to thank charlie doering , tammy kolda , and raj rao nadakuditi for useful conversations and lada adamic for providing the data for the network of political blogs .
this work was funded in part by the national science foundation under grant dms1107796 and by the air force office of scientific research ( afosr ) and the defense advanced research projects agency ( darpa ) under grant fa95501210432 . | many methods have been proposed for community detection in networks . some of the most promising are methods based on statistical inference , which rest on solid mathematical foundations and return excellent results in practice . in this paper
we show that two of the most widely used inference methods can be mapped directly onto versions of the standard minimum - cut graph partitioning problem , which allows us to apply any of the many well - understood partitioning algorithms to the solution of community detection problems .
we illustrate the approach by adapting the laplacian spectral partitioning method to perform community inference , testing the resulting algorithm on a range of examples , including computer - generated and real - world networks . both the quality of the results and the running time rival the best previous methods . |
a link @xmath0 in a homology @xmath1-sphere is called a _ graph link _ if its exterior is a graph manifold .
for example , a link in @xmath2 is a graph link if and only if it is _ solvable _ , that is , if it can be constructed by iterated cabling and connected sum operations from the unknot .
this class of links is particularly interesting , since all the links that arise in complex algebraic geometry are of this type .
eisenbud and neumann @xcite gave a classification of graph links by means of decorated trees called _
splice diagrams _ ; they also found a formula for the multivariable alexander polynomial @xmath3 of graph links in terms of the splice diagram ( see ( [ thm : multi - alex ] ) below ) .
since the conway potential function is determined up to a sign by @xmath3 , the only problem is to compute this sign . in 1999
, walter neumann @xcite succeeded in computing the conway polynomial ( that is : the one variable potential function ) of any fibered solvable link . in this article , neumann proves several formulas for fibered solvable links , and asks whether these equalities still hold for any graph link . in this paper , we compute the multivariable potential function of any ( fibered or non - fibered ) graph link in a homology sphere ( theorem [ theorem : cpf ] ) .
as a consequence , we give a positive answer to neumann s questions ( corollaries [ cor:1 ] , [ cor:2 ] and [ cor:3 ] ) .
in 1970 , conway @xcite introduced a new invariant of links called the _ potential function_. given an oriented ordered link @xmath4 in @xmath2 , its potential function is a well defined rational function @xmath5 which satisfies @xmath6 where @xmath7 means equal up to a multiplication by @xmath8 .
thus , this invariant is basically the multivariable alexander polynomial without the ambiguity concerning multiplication by units of @xmath9 $ ] . as a particular case of the potential function ,
conway defined what was later called the _
conway polynomial _ of a non - ordered link @xmath0 .
it is given by @xmath10 unfortunately , conway s paper contains neither a precise definition of the potential function , nor a proof of its uniqueness . in 1981
, kauffman @xcite found a very simple geometric construction of the conway polynomial , namely @xmath11 where @xmath12 is any seifert matrix of the link @xmath0 and @xmath13 the transpose of @xmath12 .
finally , in 1983 , hartley @xcite gave a definition of the multivariable potential function @xmath14 for any ordered oriented link in @xmath2 .
this definition was later extended by turaev @xcite to links in a @xmath15-homology @xmath1-sphere , and by boyer and lines @xcite to links in a @xmath16-homology @xmath1-sphere .
let us now state several useful properties of the potential function ; we refer to @xcite for the proofs . given an oriented ordered link @xmath4 in a @xmath15-homology sphere , there exists a well defined invariant @xmath14 related to the multivariable alexander polynomial @xmath3 of @xmath0 by the equality ( [ equ : alex ] ) .
furthermore , @xmath14 satisfies the symmetry formula @xmath17 also , if @xmath18 , where @xmath19 denotes @xmath20 with the opposite orientation , then @xmath21 finally , if @xmath22 , we have the following _ torres formula _ : @xmath23 where @xmath24 stands for the linking number @xmath25 .
eisenbud and neumann @xcite gave a classification of graph links using splice diagrams .
following neumann @xcite , we will not recall the whole construction here , just several important features of these combinatorial objects .
a splice diagram @xmath26 for a graph link @xmath0 is a tree decorated as follows : some of its valency one vertices are drawn as arrowheads , and correspond to the components of @xmath0 ; the arrowheads have weight @xmath27 or @xmath28 ( depending on whether the corresponding component of @xmath0 has `` intrinsic '' orientation or not ) ; each edge has an integer weight at any end where it meets a _ node _ ( vertex of valency greater than one ) , and all the edge weights around a node are pairwise coprime . given a non - arrowhead vertex @xmath29 of @xmath26 , there is a so - called _ virtual component _
@xmath30 : this is the additional link component that would correspond to a single arrow at @xmath29 with edge weight @xmath31 .
it is very easy , given two vertices @xmath29 and @xmath32 , to compute the linking number @xmath33 of their corresponding components ( virtual or genuine ) : if @xmath34 denotes the shortest path in @xmath26 joining @xmath29 and @xmath32 ( including @xmath29 and @xmath32 ) , then @xmath33 is the product of the edge weights adjacent to but not on @xmath34 , and of the possible arrowhead weights .
let us now recall eisenbud and neumann s formula for the multivariable alexander polynomial of a graph link ( * ? ? ?
* theorem 12.1 ) . if @xmath4 is a graph link given by a splice diagram @xmath26 , its multivariable alexander polynomial is equal to @xmath35 where the product is over all non - arrowhead vertices @xmath29 of @xmath36 , @xmath37 is the valency of the vertex @xmath29 , and @xmath38 denotes the linking number of @xmath39 with @xmath40 . in this equation ,
the terms of the form @xmath41 should be formally cancelled against each other before being set equal to zero . in proving this theorem , eisenbud and neumann
also show another remarkable result ( * ? ? ?
* theorem 12.2 ) .
let us call an @xmath42-component link _ algebraically split _ if , after possible renumbering , there is an index @xmath43 such that @xmath44 whenever @xmath45 .
the theorem says that if @xmath0 is a graph link , then @xmath46 this is a very striking property of graph links .
for example , it implies that the alexander polynomial @xmath3 of a @xmath47-component graph link is zero if and only if the linking number @xmath48 of the components is zero . for general @xmath47-component links
@xmath0 , if @xmath3 vanishes , then @xmath49 ( by the torres formula ) .
but the converse is false : the whitehead link has alexander polynomial @xmath50 , although @xmath49 . as a matter of fact
, it is still an open question how to characterize geometrically @xmath47-component links with vanishing alexander polynomial ( see ( * ? ? ?
* problem 16 ) ) . before stating and proving our results ,
let us finally recall very briefly neumann s argument @xcite for fibered solvable links .
let @xmath0 be an oriented graph link given by a splice diagram @xmath26 . using the value of the one variable alexander polynomial @xmath3 of @xmath0 , together with the equality @xmath51 , where @xmath52 is the degree of @xmath3
, we obtain the formula @xmath53 where the product is over all non - arrowhead vertices of @xmath26 , @xmath37 is the valency of the vertex @xmath29 , and @xmath54 denotes the linking number of @xmath0 with @xmath40 .
thus , the only issue is the determination of the sign of @xmath55 .
if @xmath0 is a fibered link , its seifert matrix @xmath12 is unimodular .
therefore , the leading coefficient of @xmath56 is given by @xmath57 .
for fibered links , lee rudolph defined an integer invariant @xmath58 called the _ enhanced milnor number_. if the ambient sphere is @xmath2 , this number is known to satisfy the following formula ( see @xcite ) : @xmath59 neumann proves that @xmath60 , where @xmath61 is the number of @xmath62-weighted arrowheads and @xmath63 the number of non - arrowhead vertices @xmath29 for which @xmath54 is negative and @xmath37 is odd .
this leads to @xmath64 giving the formula @xmath65 these results lead neumann to the following questions : is the formula @xmath66 true for any graph link in a homology sphere ? does the equality @xmath67 hold for fibered graph links in a homology sphere ? is the equation @xmath68 still valid for fibered graph links in a homology sphere ?
corollaries [ cor:1 ] , [ cor:2 ] and [ cor:3 ] provide an affirmative answer to these three questions .
we are now ready to state and prove our results . using ( [ thm : multi - alex ] ) along with the equations ( [ equ : alex ] ) and ( [ equ : sym ] ) , it is easy to compute the conway potential function up to a sign .
by ( [ thm : multi - alex ] ) and equation ( [ equ : alex ] ) , @xmath5 is equal to @xmath71 for some integers @xmath72 and some sign @xmath73 . the symmetry formula ( [ equ : sym ] ) implies that @xmath74 , giving the lemma .
let us use kauffman s construction of the reduced potential function @xmath79 where @xmath12 is any seifert matrix for @xmath75 .
it is easy to check that the seifert matrix for a graph knot @xmath75 is equal to the seifert matrix for the knot @xmath80 obtained from @xmath75 by deleting every non - fibered splice component .
( in other words : if @xmath26 is a splice diagram for @xmath75 , a splice diagram @xmath81 for @xmath80 is obtained from @xmath26 by deleting every vertex @xmath29 such that @xmath82 . )
therefore , we can assume that @xmath75 is fibered .
in that case , @xmath12 is unimodular . in particular
, @xmath57 is non - zero , so that @xmath57 is the leading coefficient of @xmath83 .
there is an explicit computation of @xmath12 for graph knots in @xcite ; it can be used to check the lemma .
details can be found in @xcite .
let us associate to @xmath0 a graph @xmath86 as follows : the vertices of @xmath86 correspond to the components of @xmath0 , and two vertices are linked with an edge if the linking number of the corresponding components is not equal to zero .
clearly , @xmath0 is algebraically split if and only if @xmath86 is not connected . given a vertex @xmath29 of a graph @xmath87 ,
let us denote by @xmath88 the subgraph obtained by deleting the vertex @xmath29 and every edge adjacent to @xmath29 .
we are left with the proof of the following assertion : given a graph @xmath87 , if @xmath88 is not connected for any vertex @xmath29 of @xmath87 , then @xmath87 is not connected .
in other words : if @xmath87 is a connected graph , there exists a vertex @xmath29 such that @xmath88 is connected .
this last statement is very easy to prove : given @xmath87 a connected graph , let @xmath89 be a maximal subtree of @xmath87 . since @xmath89 is a tree , it has at least one vertex @xmath29 of degree one . then , @xmath90 is connected , as well as @xmath88 . [ theorem : cpf ] let @xmath0 be a graph link with @xmath42 components given by a splice diagram @xmath26 .
then , its conway potential function is equal to @xmath91 where the product is over all non - arrowhead vertices @xmath29 of @xmath36 , @xmath37 is the valency of the vertex @xmath29 , @xmath92 , and @xmath61 is equal to the number of @xmath62-weighted arrowheads . by formula ( [ equ : inv ] )
, it may be assumed that all the arrowheads have weight @xmath76 . using the notation of lemma [ lemma : uptosign ]
, we have to check that if @xmath93 , then @xmath94 .
let us do this by induction on @xmath95 .
the case @xmath96 is settled by lemma [ lemma : knot ] .
let us fix some @xmath97 , and assume that @xmath98 for all graph link @xmath99 with @xmath100 arrowheads , all @xmath76-weighted .
let @xmath4 be a graph link with @xmath93 . if @xmath101 for all @xmath85 , then ( by ( [ equ : alex ] ) and ( [ thm : split ] ) ) @xmath84 is algebraically split for all @xmath85 .
by lemma [ lemma : split ] , it follows that @xmath0 is algebraically split , so that @xmath102 . in this case
, there is nothing to prove .
therefore , it may be assumed ( after possible renumbering ) that @xmath103 .
let us note @xmath104 .
if @xmath105 , then @xmath106 is algebraically split , so @xmath102 . hence , we can assume without loss of generality that @xmath107 .
to summarize , it may be assumed that @xmath108 the key ingredient of the induction step is the torres formula ( [ equ : torres ] ) . by lemma [ lemma : uptosign ]
, we have @xmath109 on the other hand , by ( [ equ : torres ] ) , @xmath110 which , by induction , is equal to @xmath111 now , the set of non - arrowhead vertices @xmath112 of @xmath99 is equal to @xmath113 , where @xmath114 is the vertex corresponding to the component @xmath20
. therefore , @xmath115 , @xmath116 , and we have the equality @xmath117 since @xmath118 which is assumed to be non - zero , @xmath70 is equal to @xmath27 .
if @xmath0 is a fibered link , @xmath12 is unimodular , so @xmath57 is the leading coefficient of @xmath83 . by corollary [ cor:1 ] , @xmath57 is the leading coefficient of @xmath122 , which is clearly @xmath123 . by theorem 6.1 and @xmath12510 in @xcite
, we have the equality @xmath126 where the notations are as in @xcite . by corollary [ cor:2 ]
, it remains to show that @xmath127 the argument is exactly as in @xcite . | we give a closed formula for the multivariable conway potential function of any graph link in a homology sphere . as corollaries , we answer three questions by walter neumann @xcite about graph links . |
galaxy clusters are of great interest in cosmology as they are the largest and most recently formed structures in the cosmological hierarchy .
the clustering and number density evolution of clusters are sensitive to both the growth function and the expansion history of the universe .
clusters contain a representative sample of baryons and dark matter and are thus also fascinating laboratories in which to study the influence of baryonic physics on the formation of large - scale structure , substructure , and its effect on the gravitational potential and dark matter halo shape .
statistical measures such as the mass function , the rate of structure growth , and the clustering of large structures , are fundamental predictions of cosmological models .
cluster observations are therefore expected to provide some of the most important constraints on fundamental cosmology and astrophysics ( see e.g. @xcite for a recent review ) .
cluster formation histories are frequently punctuated by large jumps in mass from major mergers ( e.g. @xcite ) .
these mergers are one of the primary mechanisms for the buildup of mass in clusters and superclusters . in the standard paradigm @xcite , observable properties such as the degree of spatial clustering depend _ only _ on the cluster mass .
however , recent theoretical studies indicate that many cluster observables , such as spatial clustering , concentration , galaxy velocity dispersion , gas distribution and its attendant observables such as x - ray emissions and sz decrement , depend on the cluster s formation time , mass accretion history , large scale environment ( collectively referred to as `` assembly bias '' ) @xcite . in addtion
, there is a dependence on recent merger history ( `` merger bias '' ) @xcite .
in addition , recent studies have claimed observational detection of assembly bias , though with mixed results @xcite .
theory can predict cluster properties as a function of their mass , which is dominated by dark matter and thus can not be directly measured . since all methods of observing clusters are sensitive to the effects of assembly / merger bias , it is necessary to develop a more detailed understanding of the mechanisms of structure formation . specifying a correlation function and mass function , and the evolution of these quantities
, may not be sufficient to connect theory to observation .
understanding cluster merger properties is therefore crucial to utilize these objects as probes of cosmology .
we focus primarily on galaxy cluster mergers .
@xcite used a large simulation volume to probe high - mass halos with good statistics .
they found that halos of mass @xmath2 that have undergone a recent ( within 1 gyr or less ) major merger or large mass gain exhibit an enhancement in their spatial clustering of up to @xmath3@xmath4 on scales of @xmath5@xmath6mpc compared with the entire halo population at the same mass .
they noted that this merger bias persists for the redshift range @xmath0 , and that the bias increases with larger merger mass ratios or shorter merger timescales .
if this merger bias remains unaccounted for , then mass - observable relations that connect theory with observation calibrated on the basis of clustering may be suspect
. moreover , the gas properties of a recently merged halo cluster can be quite different from the general population .
this can be mitigated by selecting `` relaxed '' halos ( presumably those that have not had recent major mergers ) when calibrating observables .
however , the scaling relations from this selected sample may be biased with respect to that from the overall cluster population . in addition , if a merger accidentally entered the `` clean '' sample ( e.g. because it occurred along the line of sight ) , it could substantially bias the calibration , and hence the ensuing mass determination .
it is thus important to predict the fraction of halos which have had recent merger activity . observing cluster mergers as a function of redshift ,
studying the merger bias , and correlating mergers with other tracers of density and environment will also shed light on the nature of structure formation .
a number of authors ( e.g. @xcite ) have studied the statistics of galaxy merging using close pairs as a proxy for mergers . in this work
we extend those analyses to the cluster scale .
it is important to quantify the conditions for which the pair proxy assumption is valid .
@xcite used simulations to examine the utility of galaxy pairs in measuring the redshift evolution of halo merger rates out to @xmath7 .
they tracked the formation and evolution `` host '' halos of dark matter , as well as self - bound density peaks within host halos , i.e. `` subhalos , '' a term which includes both satellite objects and the central host itself . using the assumption that galaxies are found at the centers of subhalos , they found that galaxy pair counts can probe the rates at which satellite galaxies merge with the central galaxy , i.e. _ subhalo _ mergers .
these pairs can also constrain the galaxy halo occupation distribution , i.e. the statistical number of galaxies found within a host halo of a given mass . however , galaxy pairs can not be used to measure the merger rates of separate massive _ host _ halos .
the goal of this paper is to investigate whether close spatial pairs of halo objects ( galaxy clusters ) can be used to observationally study the properties of mergers .
identifying a cluster merger in an observation is not straightforward because it relies principally on the object appearing morphologically `` disturbed '' @xcite .
furthermore , it is difficult to estimate the completeness and the contamination of such an observed merger population .
if close pairs prove to be a sufficient proxy for mergers , they would provide a clean tool to study merger statistics .
we concentrate primarily on massive halos ( @xmath8 ) , corresponding to galaxy groups and clusters , in contrast to previous work which has focused on pairs as a proxy for galaxy mergers .
however , our results also apply to any class of object that singly occupies a halo , particularly field galaxies at high redshift . in [ sec :
sim ] we describe the simulations and halo catalogs used in our study . in
[ sec : close ] we present statistics for merger rates as a function of halo pair masses , pair separations , and merger time interval to quantify the extent to which close pairs can be used as proxies for cluster mergers in the best - case scenario where halo spatial positions are known accurately . in
[ sec : redspace ] we consider observational complexities such as scatter in the halo mass , redshift space distortions , and redshift space errors . in
[ sec : kernel ] we present a mass - dependent fit to the merger kernel and examine how large - scale density impacts the efficiency of pair merging . in
[ sec : xipairs ] we discuss the prospects for studying merger bias using close cluster pairs .
we conclude in [ sec : conclusions ] .
our study is conducted using two high - resolution n - body simulations of cold dark matter in a flat @xmath9cdm cosmology with parameter values @xmath10 , @xmath11 , @xmath12 , @xmath13 and @xmath14 .
our simulations employ the hot code @xcite in a ( @xmath15 and a ( @xmath16 volume with periodic boundary conditions , using a plummer softening length of @xmath17 .
gaussian initial conditions were randomly generated for the @xmath18 particles of mass @xmath19 ( smaller simulation ) and @xmath20 ( larger simulation ) at an initial redshift of @xmath21 .
simulation outputs were stored in intervals of @xmath22 gyr between redshifts @xmath23 and @xmath24 , with the last interval of each simulation being shorter : @xmath25 gyr ( smaller simulation ) and @xmath26 gyr ( larger simulation ) .
all time intervals cited below represent the total time elapsed between two simulation outputs .
thus a merger timescale of @xmath27 gyr indicates that two separate halos have merged within 1 gyr , an upper limit to the actual time to merger .
we generate a halo catalog for each output using the friends - of - friends ( fof ) algorithm @xcite with a linking length of @xmath28 of the mean interparticle spacing .
these groups correspond to a density threshold of @xmath29 and enclose primarily virialized material . in this work
we will quote fof masses , which are about 20% smaller than `` virial '' masses ( corresponding roughly to fof masses with @xmath30 , which are more commonly found in the literature ; see @xcite for more details ) .
since we are examining mergers , we use a smaller linking length to decrease contamination that arises from close neighboring halos being bridged by a narrow string of particles ( although using a larger linking length of @xmath30 changes our results by only a few percent , see fig . [
fig : fmergemass ] ) .
the halo catalogs of the smaller simulation include all halos of mass greater than @xmath31 ( @xmath32 particles ) , though in our study we are concerned primarily with halos of mass @xmath33 , of which there are around @xmath34 at @xmath24 .
our larger ( @xmath16 but less resolved simulation catalogs include all halos of mass greater than @xmath35 ( @xmath32 particles ) .
we do not consider substructure within host halos .
merger trees were constructed from the set of halo catalogs by specifying a parent - child relationship , where a `` parent '' is any halo that contributed mass to a halo at a later time , i.e. a `` child . ''
we define a parent contributing more than half of its mass as a `` progenitor . '' under this restriction , a progenitor can never have more than one child .
we define a merger as a child having more than one progenitor . in cases where we select on a child that had more than two progenitors ,
we apply the two body approximation by considering only the two progenitors that contributed the most mass to the child . using instead the two most massive progenitors , as opposed to the two most mass - contributing progenitors ,
changes the progenitor identification in less than 1% of all mergers that we consider .
in addition to this progenitor merger tree , a list of all contributing parents ( not just progenitors ) is stored for each child .
this flexible storage of parent data allows us to study two body mergers , mergers with more than two progenitors , or mass accretion including all parents .
however , when considering short timescales ( @xmath36 gyr ) , the two body criterion is a good approximation ( see @xcite for more details ) .
all errors cited are @xmath37 errors derived from dividing the simulations into 8 octants and computing the dispersion of the quantity of interest within each octant . since we probe scales much smaller than the octants , we treat them as uncorrelated volumes .
to examine whether close pairs of galaxy clusters form reliable predictors of mergers , we first extend the work of @xcite on halo mergers , using the same simulation and halo catalogs ( described in [ sec : sim ] ) .
we identify child halos of mass @xmath38 that are products of recent mergers with progenitor masses @xmath39 and @xmath40 , where @xmath41 ( major mergers ) .
we explore the distribution of progenitor separations in fig .
[ fig : parsepz ] , which shows the number density of progenitors as a function of binned progenitor separation . for time intervals of @xmath22
gyr , there is a characteristic comoving progenitor separation ( @xmath42 ) that does not evolve significantly with redshift . without the influence of gravitational attraction , halos with typical velocities of @xmath43 will travel @xmath22 mpc within @xmath22 gyr .
furthermore , the number density of mergers evolves only weakly with redshift .
these factors suggest that close pairs might be a reasonable proxy for mergers .
these results represent a post - diction , where we know in advance which halos will merge .
we now investigate whether close pairs of objects can reliably be used to _ predict _ mergers .
the following subsections investigate how frequently pairs will merge as a function of pair mass , ( comoving ) separation , redshift , and merger time interval .
we consider a given pair to have merged if both halos are progenitors of the same child at a later timestep .
since observations of galaxy clusters are generally limited by a threshold luminosity , we identify pairs of halos whose individual halo masses are above a given mass cut .
we first examine the pair - merger hypothesis ( that close pairs are merger proxies ) for the best - case scenario in which halo positions and masses are known with complete accuracy ( as in our simulations ) .
this will provide a firm upper limit to the utility of pairs in predicting mergers .
then in [ sec : redspace ] we consider observational complexities such as scatter in the halo mass , redshift space distortions , and redshift space errors , which significantly degrade the signal .
we select halos at @xmath44 and consider mergers to @xmath24 ( @xmath45 gyr ) .
figure [ fig : mergerstat_zis0 ] ( top ) shows the number density of pairs as a function of pair separation , @xmath46 , along with the number densities of pairs that merged ( solid ) , @xmath47 , and did not merge ( dashed ) within the time interval .
the number density of pairs terminates at small separations because of halo exclusion , i.e. massive halos have finite radii .
the upper set of curves are halos of mass @xmath33 and the lower have @xmath48 ; the results are qualitatively similar .
there is a limited range of separations larger than the halo exclusion limit in which the majority of pairs merge .
also plotted ( center ) are the same number densities , but as an integrated function of separation , demonstrating the relationship for pairs _ within _ the given separation @xmath49 . in @xmath25
gyr at @xmath24 , massive halo pairs will only merge with certainty for separations @xmath50mpc .
figure [ fig : mergerstat_zis0 ] ( bottom ) also shows the fraction of pairs that merge _ within _ the given separation , @xmath51 ( falling curves ) .
this fraction is the `` efficiency '' of the close - pair method in identifying merger candidates , since it shows the likelihood that a pair within the given separation will merge .
it can also be thought of as a measure of contamination of the candidate sample , because its difference from @xmath22 identifies the fraction pairs that do not merge .
the rising curve shows the fraction of mergers over the _ total _
number density of mergers , @xmath52 .
this can be interpreted as the merger `` completeness , '' as it shows the fraction of the total number of mergers found by identifying pairs within the given separation .
figure [ fig : mergerstat_zis0 ] confirms the intuitive result that because of greater accelerations , larger halos are able to merge from larger separations , which holds at all redshifts and time intervals ( see fig . [
fig : fmergemass ] ) . if close pairs were ideal predictors of mergers , the rising and falling curves of fig .
[ fig : mergerstat_zis0 ] ( bottom ) would be step functions , crossing near a fraction ( @xmath53-value ) of @xmath22 and demonstrating a clear dichotomy between those pairs that merge and those that do not .
however , because of dynamical effects halos do not simply accelerate toward one another from rest ; they merge with neighbors from a broad distribution of separations .
indeed , the effects of local environment can hinder the merging of close pairs that would ordinarily occur via non - linear two - body gravitational interaction ( see [ sec : kernel ] for more details ) .
the intersection of the efficiency and completeness curves represents an easily identifiable pair separation that compromises between maximizing completeness while minimizing contamination . at this intersection
the fraction of pairs that merge , @xmath51 , equals the fraction of total mergers , @xmath52 , although the relevant populations are different .
if this intersection occurs at a fraction close to @xmath22 , most pairs within the separation will merge , and those pairs will represent the majority of all mergers that occur in the time interval . however , if this fraction is low , pairs within the separation are not a representative indicator of mergers . since the intersection of these two curves occurs at a fraction of @xmath54 for massive pair mergers within @xmath55 gyr at @xmath24 , we conclude that pairs at these redshifts and masses can not be used to reliably predict mergers .
we next examine whether the pair - merger hypothesis fares better at high redshift .
we consider a longer merger timescale of @xmath27 gyr for massive halo pairs at @xmath56 and @xmath57 .
figure [ fig : fmergez ] ( top ) shows the pair merger fractions as a function of ( comoving ) separations .
the pair merger efficiency ( falling curves ) increases significantly with redshift , while the merger completeness ( rising curve ) remains approximately invariant . for a given merger timescale ,
mergers come from pairs of essentially the same ( comoving ) separation , regardless of redshift ( see fig .
[ fig : parsepz ] ) .
however , since gravitational interactions are governed by physical ( not comoving ) separations , fig .
[ fig : fmergez ] ( bottom ) shows the same merger fractions as a function of physical separation .
the merger efficiency trend is reversed : pairs within a given physical separation are more likely to merge at low redshift , and mergers come from pairs of much larger physical separations at low redshift .
the intersection of the efficiency and completeness curves occurs at a higher fraction at high redshift than at low redshift ( 75% vs 60% ) , indicating that , for a fixed merger timescale , massive halo pairs provide a slightly better proxy for merger rates at higher redshift .
while it appears that cluster - mass halo pairs might provide a reasonable proxy for mergers at @xmath58 , we next examine whether these results are robust as a function of halo mass .
figure [ fig : fmergemass ] ( top ) shows the pair merger fractions for halo masses down to @xmath59 at @xmath58 , where the pair separation is shown as a multiple of the minimum mass cut virial radius ( @xmath60 ) .
figure [ fig : fmergemass ] ( middle ) shows the same but for higher halo masses and a shorter time interval at @xmath61 .
while pairs of higher mass halos merge from larger physical pair separations , when scaled by the halo virial radius we find that the pair merger efficiency and completeness exhibit a nearly universal relation as a function of halo mass .
this implies that any of our results in [ sec : close ] for a given mass can be approximately scaled to those of another mass .
this universal merger relation also implies that the total merger fraction can be estimated by noticing that the following relation is satisfied at the crossing point of the curves : @xmath62 .
for example , from the middle panel of fig .
[ fig : fmergemass ] we find that @xmath63 when evaluated at a scaled pair separation of @xmath64 .
one can thus estimate the merger rate by counting pairs interior to @xmath64 .
this is a potentially powerful result since it is approximately invariant for all mass - scales .
however , this radius is a function of the timescale considered , and is likely to depend on the cosmological parameters .
furthermore , although their number densities agree , the pair population within @xmath64 does not directly corespond to the total merger population .
this makes it impossible to use the pair population to directly study any other properties of mergers ( e.g. , clustering ) .
redshift space distortions will significantly undermine the utility of these results ( see[sec : redspace ] ) .
the only strong deviation from this nearly universal mass relation occurs for merger completeness of @xmath65 halos ( top ) , which results in an intersection of merger efficiency and completeness at a lower fraction ( 1015% decrease ) for lower - mass halos , a trend which we find does not depend strongly on redshift or merger timescale .
the use of pairs as proxies for merger rates becomes increasingly unreliable as we approach massive galaxy - size halos , an important result for massive _ galaxy _ mergers at high - redshift , where galaxies are found primarily in distinct host halos .
these results augment those of @xcite , who found that the evolution of close galaxy pairs can not be used to measure the host halo merger rate .
specifically , they found that the host halo merger rate evolves as @xmath66 , while the number of close galaxy pairs evolves little with redshift .
this arises because , at low redshift , the merger rate of host halos is low , but there are multiple galaxy pairs merging _ within _ massive host halos . at high
redshift , the host halo merger rate is high , but the number density of halos massive enough to host more than one bright galaxy is low .
our results imply that even when considering major mergers of massive galaxies at high redshift , where each halo hosts a single bright galaxy , the close pair population will not reliably trace the merger population .
we have also compared two different linking lengths , to explore the dependence of our results on the fof process . in fig .
[ fig : fmergemass ] ( bottom ) , we show the results using a linking length of @xmath67 . since changing the linking length changes the mass of a given halo , we have scaled the mass threshold for @xmath67 to match the number densities of the @xmath28 sample , thereby probing the same halo population . we have also scaled the halo separations by @xmath68 . in the case of @xmath67 , the intersection of the efficiency and completeness curves
is shifted outward by @xmath69 , which is not surprising since a halo found using @xmath67 is expected to be @xmath70 larger than one found using @xmath28 . from the figure
we see that changing the linking length results in a change of only a few percent in the merger fraction at the intersection of the efficiency and completeness curves .
thus changing the linking length does not qualitatively alter any of our results .
the weak dependence on the linking length suggests that for these mergers , the fof procedure does not give rise to significant artificial bridging of nearby halos , and our results would not differ significantly from a similar analysis performed using a spherical overdensity halo finder @xcite .
pair merging is also strongly dependent on the choice of merger timescale .
figure [ fig : fmergezishalf ] shows that the merger efficiency ( falling curves ) increases with timescale , i.e. pairs are more likely to merge when one considers longer time intervals . when considering mergers across @xmath5 gyr , a significant fraction ( 80% ) of pairs with separations @xmath71mpc will merge .
however , the merger completeness within a given separation ( rising curves ) decreases with timescale , so one must consider pairs at larger separation to capture all mergers .
interestingly , the trade off between efficiency and completeness is only weakly dependent on the timescale , and we find similar results even for massive halo pairs at @xmath72 merging to @xmath73 ( @xmath74 gyrs ) . no matter which merger timescale are considered , the utility of pairs as proxies for mergers remains limited . in all redshift , mass , and temporal regimes we consider , close halo pairs
do not provide a robust predictor of overall merger rates .
we conclude that , while pairs at small separations can reliably predict mergers , these constitute a small fraction of the total merger population .
while our results could in principle be used to calibrate the merger fraction as a function of halo separation , close pairs do not provide a _ self - consistent _ probe of the merger population independent of our theoretical predictions ( which are contingent upon having simluated the correct cosmology ) . _
thus , even in the best - case scenario of complete knowledge of mass and position , measurements of galaxy cluster pairs can not be used to measure cluster merger rates .
the same is true for any any class of object that singly occupies a halo , particularly in using field galaxy pairs at high redshift to probe galaxy mergers . _
the situation becomes more complicated when we consider redshift space distortions , redshift space errors , and scatter in the estimated mass of the halos . to identify the impact of imprecise determination of dark matter halo masses
, we have recomputed several of the statistics of [ sec : close ] after introducing an rms scatter of @xmath75 , where @xmath76 is the threshold for detecting the pairs in the mock observation .
the scatter in the mass causes some halos to fall out of the sample and others to enter it , resulting in a different population of halos near the threshold . because the mass function is steep , many more low - mass halos enter the sample than high - mass halos leave it .
thus there are both more pairs to consider as merger candidates and more actual mergers between members of the observed sample .
for example , if @xmath77 , the number of pairs increases more than the number of mergers , resulting in a decrease in @xmath78 of @xmath79% for pairs separated by @xmath80mpc over @xmath81 ( @xmath82 gyrs ) . the effect is similar for other thresholds , redshifts , and intervals .
scatter in the mass has a more pronounced impact at high values of @xmath76 where the mass function is steeper .
the trend with redshift is derivative of the evolution of the mass function : at a fixed value of @xmath76 there is more sensitivity to scatter at higher redshifts where the mass function steepens .
there is not a significant trend with merger timescale @xmath83 .
the overall impact of imprecise mass measurements is to degrade the range in separation over which pairs might be considered useful proxies for mergers , but it is not a significant effect .
in contrast , redshift space distortions from doppler shift due to the halo peculiar velocities @xmath84 are catastrophic for the pair - merger hypothesis . when pairs are identified in redshift space , virtually none of them merge . in redshift space ,
the line of sight component is shifted relative to its real position by an amount @xmath85 for typical halo velocities in our simulation , this shift amounts to a few @xmath86mpc .
figure [ fig : redbad ] illustrates that this distortion makes many close pairs appear to be highly separated and pairs that are highly separated to appear close . at redshift
space separations less than @xmath87mpc , fewer than 5% of merger candidates actually merge in @xmath88 gyr at @xmath24 .
even at @xmath89 where redshift space distortions are less severe , only @xmath9010% of candidates merge over @xmath91gyr .
the situation improves modestly for longer time intervals , with @xmath9060% of pairs merging over @xmath92gyr , but this falls far short of the fraction in real space .
figure [ fig : redbad ] shows a histogram of the difference between a pair s separation in redshift space ( dashed line ) , as would be done in an observation , and its true separation in configuration space ( solid lines ) at @xmath44 .
merger candidates identified in real and redshift space are widely disjoint sets .
the three sets of curves ( colors ) are bins of separation , as annotated . for the population of merger candidates identified in redshift space ( dashed lines ) ,
the region greater than zero shows pairs for which the real separation was smaller than the measured separation , while the negative region shows pairs that appeared to be close in redshift space but were actually separated by several @xmath86mpc .
the former set might be likely to merge , but the latter set is highly unlikely to do so .
for the actual population of close pairs ( solid lines ) , the region greater than zero shows pairs whose redshift space separations were greater than their real separations , and which were therefore likely removed from the sample of merger candidates .
the reason for the asymmetry between the pair population that left the sample ( solid , right side ) and the population that entered the sample and replaced it ( dashed , left side ) is that pairs are most often found in overdense regions , toward which many other halos are streaming .
the impact is that the pair population identified in redshift space is larger and has a broad distribution of physical separations ; virtually none of the `` merger candidates '' we identified actually merge .
this result is robust for all redshifts , merger intervals , and halo masses .
to a degree , this result also applies to the relationship between galaxy pairs and galaxy mergers . observing galaxy pairs to study the merger rates of singly occupied host halos at high redshift
will be similarly affected by redshift space distortions , as will using galaxy pairs to deduce the merger rates of multiply occupied host halos at lower redshift . however , the impact of redshift space distortions on using galaxy pairs to deduce the galaxy - galaxy merger rates within a single host halo may be less severe @xcite . inside a halo ,
galaxy mergers are predominately between a satellite galaxy and the central galaxy , and merger dynamics inside a host halo are influenced by dynamical friction , especially in group - mass halos and when the galaxy masses are comparable .
pair counts , even in redshift space , provide a rough proxy to the halo mass and will therefore correlate well with the merger rate within the halo .
the impact of redshift space errors is similar to that of redshift space distortions , except that , unlike with the velocities , the errors are not correlated with the high density regions .
redshift space errors that result in more than a few @xmath86mpc shift in the apparent position have a similar impact to redshift space distortions .
since spatial information is a weak probe of halo merger statistics , we now turn to statistically describing the merger rate via the halo mass function . for any population of objects that are built up by binary mergers of smaller constituents , the rate at which the number of objects of a given mass changes can be described by the smoluchowski coagulation equation , which has been applied to the evolution of dark matter halos @xcite .
the abundance of halos at a given mass is increased through the creation of such halos by mergers of smaller halos , and decreased as these halos merge into more massive ones .
the rate of change of the number density of halos of a given mass can then be determined by knowing the number density of halos at all masses , i.e. the mass function , and the proper merger kernel to relate the mass function to a merger rate .
historically , it has been difficult to study the coagulation of cluster - mass halos through simulation because these events are rare . however , with our large simulation volume the merger kernel can be computed in a statistically significant way .
we define the merger kernel @xmath93 as in @xcite ( hereafter f&k ) , via the relation @xmath94 where @xmath95 is the number density of mergers between parents of mass @xmath96 and @xmath97 in a time @xmath83 , and @xmath98 is the halo mass function . the quantity @xmath99 can be interpreted as the efficiency of merging for a pair of objects , such that the rate of mergers is a product of this efficiency with the densities of available parents .
note that in previous contexts the term efficiency refers to the ability of the close - pair technique to reliably identify merger candidates , whereas in this context the efficiency is the coefficient of the number densities when quantifying the merger rate .
the time interval @xmath83 should be sufficiently short that there is no significant evolution of the halo mass function . satisfying this requirement drives down the number of mergers in a fixed volume , making statistically significant measurements difficult
however , with our large simulation volume we measure @xmath100 on the interval @xmath101 ( @xmath55 gyr ) by classifying the two progenitors that contributed the most mass to each halo at @xmath24 into ten logarithmically spaced mass bins in the range @xmath102 .
the results are shown in fig .
[ fig : merkernel ] we find that @xmath99 follows the simple functional form @xmath103^b \ , .\ ] ] the best fit values for @xmath104 and @xmath105 , found using a linear least - squares fit to the data , are presented in table [ tab : merkernel ] ( bottom row ) .
this functional form satisfies the formal requirement that the merger kernel be symmetric in its two arguments . for comparison
, @xcite analytically found that @xmath106 when @xmath107 with @xmath108 , which is approximately true in the trans - linear regime of cluster - mass halos .
.best fit amplitude , @xmath104 , and exponent , @xmath105 , for the merger kernel @xmath109 , where @xmath110 is the sum of the 2 progenitor masses .
the first four rows are the results for each of the 4 density subdivisions , while the bottom row is the best fit for the entire simulation .
the average density and total volume of each subdivision is listed at left , and the reduced @xmath111 for each subdivision is listed at right . [ cols="^,^,^,^,^",options="header " , ] because the densities of the parent populations have been divided out , it is conceivable that the merger kernel @xmath99 is independent of the large - scale density field .
significant dependence on density would indicate that environmental effects other than the progenitor densities , such as halo velocity distributions , halo impact parameters , and tidal fields , are at play in determining the merger rate . to investigate the dependence of @xmath99 on the large - scale density field we construct a coarse ( @xmath112 ) density grid in our simulation and assign the dark matter particles to the nearest grid point , effectively smoothing the field on a scale of @xmath113mpc .
each halo at @xmath24 is also assigned to the coarse grid , and these halos are sorted in order of their large scale density environment and then divided into quartiles of density , with each quartile containing approximately the same number of halos . using the coarse grid , we compute the total volume and overdensity of each quartile in the simulation , and these appear in the first two columns of table [ tab : merkernel ] .
the mean overdensity of a quartile is defined as @xmath114 where the angular brackets denote an average of the @xmath90(@xmath115mpc)@xmath116 cells of the coarse grid . the merger kernel @xmath99 is fit for each quartile to the form of eq .
[ eqn : qform ] , using the value of @xmath105 determined from the entire simulation .
the results are summarized in table [ tab : merkernel ] .
the best fit values of @xmath104 differ by several sigma , and the improvement in the reduced @xmath117 statistic indicates that each of the individual density fits is a much better fit that a fit to the entire volume .
this indicates that there is a clear trend in @xmath99 with the large - scale density field : as the density increases the efficiency of merging for a given system mass decreases .
_ thus , the environmental effects in dense environments are hindering the merging process in comparison to less dense environments .
_ figure [ fig : merkernel ] shows @xmath99 in each of the four density environments plotted individually .
the upper panel shows @xmath99 in the highest and lowest density quartiles , while the lower panel shows the results from the inner two quartiles .
the solid line is the best fit model for the entire box , which is only a good approximation for environments close to the mean density .
the yellow shaded region results from allowing @xmath104 and @xmath105 to deviate by their @xmath37 values .
figure [ fig : merkernel ] demonstrates clearly that a large component of the dispersion in the best fit to the entire simulation originates in the density dependence of @xmath99 .
figure [ fig : merkernel ] clearly indicates that merging is less efficient in high density environments than in low density environments , but we note that with regard to our fit there is an ambiguity as to whether this implies a lower normalization @xmath104 or a shallower power law @xmath105 . given the rarity of these events at the high mass end and the limited dynamic range that is consequently driving the fit , there is large covariance between the fit values of @xmath104 and @xmath105 . we have performed the 2-parameter fit to eq . [ eqn : qform ] for each of the density quartiles individually and find that indeed @xmath104 increases slightly with density ( in contrast to the 1-parameter fit ) while @xmath105 decreases .
unfortunately , the formal errors in the parameters ( from inversion of the covariance matrix ) become so large that the four regions are statistically indistinguishable .
however , the individual fits hint at an interesting possibility : if @xmath105 decreases and @xmath104 increases with density , then the merger efficiencies cross over as a function of system mass . specifically , for lower total system masses ( @xmath118 ) , merging becomes more efficient in high density environments . in our simulation , all four @xmath99 curves cross over in the mass range @xmath119 , suggesting that galaxy scale mergers may be more efficient in denser environments .
while this evidence is extrapolated from higher mass , the trend is potentially worth further investigation given that other halo properties , e.g. clustering as a function of concentration / formation time , reverse trend from @xmath120 to @xmath121@xcite is the mass at which @xmath122 , the variance of the linear power spectrum smoothed on scale @xmath123 , equals the threshold for linear density collapse @xmath124 ; @xmath125 at @xmath24 . ] .
we note that the merger kernel @xmath99 as computed in this section is not a direct cosmological observable .
there is an observational counterpart to the merger kernel : the pair kernel , @xmath126 , computed using the number density of pairs with separations @xmath127 .
the threshold separation is calibrated from simulations to any desired tolerance for completeness or contamination , but the result is insensitive to @xmath128 . in all cases ,
there are many more pairs than actual mergers , the amplitude @xmath104 of @xmath126 is larger , and the power law @xmath105 is shallower , both by several @xmath129 . it infeasible to use pairs at any mass scale as a proxy to compute merger rates .
the trend in the merger kernel is driven by the fact that the number of mergers grows more slowly with density than the parent mass functions . by studying changes in the relationship between close pairs and mergers in regions of differing local density
, we can demonstrate that this is not universally true for all pair separations .
we proceed in an analogous manner , smoothing over @xmath113mpc and defining four density quartiles , each containing the same number of mergers .
we consider pairs merging between @xmath130 and @xmath44 ( @xmath27 gyr ) , though our results are insensitive to the time interval and redshift .
[ fig : fmerge_density ] shows the results for pairs of halos above @xmath65 ( top ) and @xmath131 ( bottom ) .
the relationship between close pairs and mergers changes with the large - scale density .
pairs within a given separation are a less complete sample of merger candidates in overdense environments than in underdense regions ( rising curves , top and bottom ) .
objects in overdense regions have higher velocities , which allow mergers to come from a broader range of progenitor separations and can extend the infall time of closer pairs by generating larger angular momenta .
figure [ fig : fmerge_density ] also indicates that , at large separations , pairs are more efficient predictors of mergers in overdense regions than in underdense regions ( falling curves ) because higher velocities in overdense regions allow pairs from larger separations to merge .
however , at small separations ( @xmath50mpc ) , the relationship between pairs and mergers depends on the pair masses . here , there are no high - mass ( @xmath132 ) halos because of halo exclusion , and pairs are more efficient predictors of mergers in _ underdense _ regions ( top )
this is likely because low - mass progenitors in overdense regions have large velocities and thus have longer infall times , e.g. from tidal fields , resulting in large angular momenta that inhibit merging .
in contrast , the smaller velocities in underdense regions enhance the probability that a very close pair will merge from simple gravitational attraction .
high - mass halos have much stronger gravitational attractions and are thus less affected by the large - scale density field . indeed , at small separations , massive halos are equally likely to merge in over- and underdense environments . finally , we note that the two methods we have presented to examine the density dependence of merging probe mergers in different ways .
the merger kernel integrates over merger pair separations , thus considering mergers as a function of progenitor pair mass sum .
alternately , the close pair merger method selects only close halo pairs that are above a given mass cut , thus integrating over mass dependence and considering mergers as a function of pair separation .
the merger kernel indicates that for all child halo masses in our simulation merging is more efficient in underdense environments .
this is corroborated by fig .
[ fig : fmerge_density ] ( top ) , which shows that for all halos in our simulation , pair merging is more likely in underdense environments for pairs within @xmath133mpc separation .
the accompanying merger completeness ( rising ) curves show that pairs within this separation constitute the majority of pair mergers in the given timescale .
figure [ fig : fmerge_density ] ( bottom ) shows that high - mass pairs are more likely to merge in overdense regions at all pair separations .
however , since mergers between two halos both with @xmath48 are extremely rare ( as opposed to mergers between a high and low mass object ) , these contribute less to the merger kernel at a given child mass .
for two halos to merge , they must have been located in close physical proximity at an earlier time .
since closely spaced halos are more likely to be found in overdense regions , recently merged halos may exhibit enhanced spatial clustering .
moreover , if recently merged halos cluster differently from the general population ( `` merger bias '' ) and this is unaccounted for , conclusions drawn about halos on the basis of their clustering could be compromised .
for example , the use of cluster self - calibration to infer cluster masses ( and in turn cosmological parameters ) requires a precise knowledge of the clustering of clusters as a function of mass @xcite .
a number of earlier studies have looked for such merger bias @xcite , with mixed results .
recently , @xcite developed an analytic model to predict the merger bias . assuming that mergers correspond to closely spaced objects at an earlier time
, they compare the clustering of close pairs to that of single objects , thus computing the pair bias as a proxy for merger bias . on scales much larger than the pair separations
, they found an enhancement of clustering for pairs of mass @xmath134 , implying that recently merged high - mass halos should exhibit a clustering bias .
we now use f&k s framework to determine if the clustering of close halo pairs of mass @xmath135 provides an accurate proxy for the clustering of mergers .
using simulations , @xcite found the most prominent merger bias ( @xmath136% ) for major mergers ( @xmath137 ) of high - mass halos ( @xmath2 ) at @xmath24 over @xmath88 gyr .
we consider a similar mass and temporal regime . to improve our statistics
we additionally use a box of eight times the volume previously used ( see [ sec : sim ] ) , which allows us to probe merger effects of more massive halos .
we use a shorter timescale of @xmath138 gyr at @xmath24 to define our merger interval to preserve the signal in the comparison population .
contrary to expectation , the merger bias _ does not _ increase with mass ; it remains a 1020% enhancement with similar statistical significance up to halos of mass @xmath139 .
in computing the pair bias , f&k define `` close '' halo pairs by demanding that the probability of finding three or more halos in a sphere of a given radius is small compared to that of finding two .
this is approximately equivalent to the restriction that the probability of finding two or more neighbors , @xmath140 , within a given separation from a halo is small compared to that of finding one , @xmath141 , which is what we will measure . for halos @xmath142 at @xmath61
, we find that @xmath143 for a ( comoving ) sphere radius of @xmath144mpc .
this separation restriction yields @xmath145 pairs per @xmath146gpc@xmath147 , out of @xmath148 halos per @xmath146gpc@xmath147 . while this is a sufficient number density of pairs for a robust correlation function measurement in our ( @xmath149gpc)@xmath116 simulation volume , [ sec : close ] demonstrated that close pairs do not reliably predict the merger population .
the analytical model of f&k predicts a significant merger bias in its application to the clustering of massive halos ( @xmath135 ) . for such objects
, it predicts a correlation function of pairs @xmath150 in terms of @xmath151 , the correlation function of individual halos : @xmath152^{4}-1.\ ] ] this is computed on scales where the underlying matter fluctuations are linear , which at @xmath24 corresponds to distances greater than @xmath153mpc . to compute the pair bias f&k use @xmath154 for a given halo mass , this can result in anomalously high pair biases .
this is because the halo correlation function is implicitly a function of the halo mass ; more massive halos are more highly clustered .
when computing the merger bias ( or its proxy , the pair bias ) the comparison population should be an ensemble of halos of the same characteristic mass as the child halo , not the parent halos . otherwise , the merger bias becomes entangled with the effect that larger halos are more clustered . to adjust for this in the analytic model we note that @xmath155 near @xmath156 . on scales
@xmath157 mpc , @xmath158 is less than @xmath22 , so this leads to a pair bias of @xmath159 .
this pair bias is still significantly larger than the 1020% merger bias seen in simulation for halos up to @xmath160 . as an alternative to an analytic approach ,
we next measure the clustering of close pairs in simulations to discover whether pair bias can predict merger bias .
we consider pairs @xmath26 gyr prior to @xmath24 . to assign a unique position to a pair of neighboring halos ,
we impose a `` couples '' restriction , namely that each member of a pair is the other s nearest neighbor .
this restriction remains robust for two body mergers , a good approximation for our time interval . within the pair separations we consider ( @xmath161mpc )
couples constitute essentially all pairs .
we select couples of halos above @xmath162 at @xmath44 and use their geometric centers to evaluate the couple correlation function .
we use couples with separations less than @xmath163mpc , 80% of which correspond to mergers in our time interval ( fig .
[ fig : fmergemass ] ) . to limit the effect of mass - dependence on the halo correlation function
, we compare this correlation function with that obtained from single halos above @xmath131 at @xmath24 .
while this ignores the effects of mass scatter in mergers , such effects remain small for this short timescale .
figure [ fig : xipairs ] shows that with a ( @xmath149gpc)@xmath116 simulation volume , no statistically significant pair bias is found .
such poor statistics arise because only a small fraction ( 15% ) of total mergers are represented by couples at such close separations . trying to increase statistics by considering couples at larger separations
is undermined by the fact that the fraction of couples that merge is a steeply decreasing function of separation .
as already mentioned , one can not consider longer merger timescales , as this permits a larger fraction of the halo population to have undergone a major merger , thereby washing out the signal in the comparison population .
we find similar results when looking at the pair bias for mass cuts from @xmath164 to @xmath165 , and thus we conclude that _ pair bias can not be used to reliably predict merger bias observationally , in simulation , or through current analytic treatment . _
cluster merger statistics may provide insight into the nature of hierarchical structure formation and the mechanisms by which the largest coherent objects in the universe form .
we use large - volume , high - resolution n - body simulations to investigate the utility of close spatial pairs of galaxy clusters as proxies for cluster mergers .
we characterize merger statistics through the merger kernel , and examine the density dependence of merger efficiency .
we highlight our conclusions as follows : * close pairs of galaxy clusters at very small separations ( @xmath166mpc ) can be used to reliably predict mergers .
however , since these constitute a small fraction of the total merger population , close pairs are not a reliable proxy for cluster merger rates .
we quantify this by measuring the efficiency and completeness of merger candidates identified via close pairs as a function of separation and find that their intersection typically occurs at a low merger fraction ( @xmath26@xmath25 ) .
* we find that close pairs are even poorer proxies for mergers between massive galaxy - sized halos .
this indicates that galaxy pairs will not provide a reliable proxy for galaxy merger rates at high redshift , where most galaxies reside in distinct halos .
* we note that the failure of close pairs as proxies for mergers indicates that determination of merger rates from spatial statistics , such as the correlation function , can not be trusted outside the physical size of a single halo . *
we examine the mass , redshift , and timescale dependence of pair mergers , finding that the pair - merger hypothesis ( that close pairs are proxies for mergers ) at a given comoving separation is most accurate at high redshift , high mass , and over long merger timescales . in our best scenarios ,
the intersection of merger efficiency and completeness is at @xmath167% , i.e. 75% of pairs within a given separation merge , and these constitute 75% of all mergers .
we also find a nearly universal relation for pair merger efficiency and completeness for different mass halos .
this relation begins to break down as we approach massive galaxy - size halos , and is compromised by redshift space distortions .
* redshift space distortions have a devastating impact on detecting close galaxy cluster pairs in surveys ; nearly all of the merger candidates identified in redshift space do not merge . although an extrapolation , we expect these results to be robust for galaxy - size halos at high redshift .
* we present the first fit from simulation to the merger kernel a means to describe halo merger rates via the halo mass function ( coagulation ) . *
the merger kernel exhibits dependence on local ( @xmath113mpc ) density .
specifically , halo merging in our high - mass regime is more efficient in underdense regions .
* pairs at large separations ( @xmath168mpc ) are more likely to merge in overdense regions . for pairs at small separations ,
low - mass halos are more likely to merge in underdense regions , while high - mass halos exhibit no environmental dependence .
* we sought to use cluster pairs to measure the merger bias by using the pair bias as a proxy for merger bias .
we extended the treatment of previous analytic work to include the fact that mergers result in mass gain ; when computing the bias of recently merged halos , the comparison population should be a set of halos of the same mass as the children , instead of the parents .
close spatial pairs that reliably merge are too rare to produce a statistically significant measure of merger bias , even in a ( @xmath149gpc)@xmath116 simulation volume . in conclusion , we have shown that close spatial pairs of galaxy clusters are of limited value as a probe of overall cluster merger rates .
we have determined the merger kernel for halo coagulation for the first time from simulations , finding that a statistical description of halo mergers is of more promise .
further work is needed to extend our parametrization of the merger kernel to lower masses and higher redshifts , and to explore whether the environmental dependence of the merger rate persists in these regimes .
we thank g. jungman and r. yan for enlightening conversations , and j. cohn , s. furlanetto , m. kamionkowski , and m. white for valuable comments on an early draft .
computational resources were provided by the lanl open supercomputing initiative and the space simulator beowulf cluster .
a.s . and a.w
were supported by nasa .
bell , e.f . ,
phleps , s. , somerville , r.s . , wolf , c. , borch , a. , & meisenheimer , k. 2006 , , 652 , 270 benson , a.j .
, kamionkowski , m. , & hassani , s.h .
2005 , , 357 , 847 berlind , a.a . ,
kazin , e. , blanton , m.r .
, pueblas , s. , scoccimarro , r. , hogg , & d.w .
2006 , , submitted ( astro - ph/0610524 ) berrier , j.c . ,
bullock , j.s . ,
barton , e.j . ,
guenther , h.d . ,
zentner , a.r . , & wechsler , r.h .
2006 , , 652 , 56 borgani , s. 2006 , astro - ph/0605575 cavaliere , a. , colafrancesco , s. , menci , n. 1992 , , 392 , 41 cohn , j.d . & white , m. 2005 , aph , 24 , 316 cole , s. & kaiser , n. 1989 , , 237 , 1127 conselice , c.j .
2006 , , 638 , 686 croton , d.j .
, gao , l. , & white s.d.m .
2006 , , 374 , 1303 davis , m. , efstathiou , g. , frenk , c.s .
, & white , s.d.m .
1985 , , 292 , 371 de propris , r. , liske , j. , driver , s.p . ,
allen , p.d .
, & cross , n.j.g .
2005 , , 130 , 1516 de propris , r. , conselice , c.j . , driver , s.p . , liske , j. , patton , d. , graham , a. , & allen , p. 2007
, , accepted ( astro - ph/0705.2528 ) efstathiou , g. , frenk , c.s . , white , s.d.m . , & davis , m. 1988 , , 235 , 715 furlanetto , s.r . & kamionkowski , m. 2006 , , 366 , 529 gao , l. , springel , v. , & white , s.d.m .
2005 , , 363 , l66 gao , l. & white , s.d.m .
2007 , , 377 , l5 gottlber , s. , kerscher .
m. , kravstov , a.v . ,
faltenbacher , a. , klypin , a. , & mller , v. 2002 , a&a , 387 , 778 hahn , o. , porciani , c. , carollo , c.m . , & dekel , a. 2007 , , 375 , 489 harker , g. , cole , s. , helly , j. , frenk , c. , & jenkins , a. 2006 , , 367 , 1039 hartley , w.g . , gaxxola , l. , pearce , f.r . ,
kay , s.t .
, & thomas , p.a . , submitted ( astro - ph/0710.3698 ) infante , l. , et al . , 2002 , , 567 , 155 jeltema , t.e . ,
hallman , e.j . ,
burns , j.o . , motl , p.m. in press ( astro - ph/0708.1518 ) jing , y.p . ,
suto , y. , & mo , h.j .
2007 , , 657 , 664 kaiser , n. 1984 , , 284 , l9 lacey , c. & cole , s. 1994 , , 271 , 676 lin , l. , et al .
2004 , , 617 , l9 maccio , a.v . ,
dutton , a.a .
, van den bosch , f.c .
, moore , b. , potter , & d. , stadel , j. 2007 , , in press ( astro - ph/0608157 ) majumdar , s. , & mohr , j.j .
2003 , , 585 , 603 masjedi , m. , et al . , 2006 , , 644 , 54 mo , h.j . , & white , s.d.m .
1996 , , 282 , 347 patton , d.r . ,
pritchet , c.j . ,
yee , h.k.c .
, ellingson , e. , & carlberg , r.g .
1997 , , 475 , 29 patton , d.r . ,
carlberg , r.g .
, marzke , r.o . ,
pritchett , c.l .
, da costa , l.n . , & pellegrini , p.s .
2000 , , 536 , 153 patton , d.r . ,
pritchet , c.j . ,
carlberg , r.g .
, marzke , r.o . ,
yee , h.k.c . , hall , p.b . , lin , h. , morris , s.l . , sawicki , m. , shepherd , c.w . , & wirth , g.d .
2002 , , 565 , 208 percival , w.j . ,
scott , d. , peacock , j.a .
, & dunlop , j.s .
2003 , , 338l , 31 poole , g.b . ,
babul , a. , mccarthy , i.g . ,
fardal , m.a . ,
bildfell , c.j .
, quinn , t. , & mahdavi , a. 2007 , , 380 , 437 rowley , d.r . ,
thomas , p.a .
, & kay , s.t .
2004 , , 352 , 508 sandvik , h.b . ,
moller , o. , lee , j. , & white , s.d.m .
2007 , , 377 , 234 scannapieco , e. & thacker , r.j .
2003 , , 590l , 69 sheth , r.k . & pitman , j. 1997 , , 289 , 66ms sheth , r.k .
& tormen , g. 1999 , , 308 , 119 sheth , r.k . & tormen , g. 2004 , , 350 , 1385 silk , j. & white s.d.m .
1978 , , 223 , l59 wang , h.y . ,
mo , h.j . ,
& jing , y.p .
2007 , , 375 , 633 warren , m. & salmon , j. 1993 , supercomputing 93 , 12 wechsler , r. h. , bullock , j.s . ,
primack , j.r . ,
kravstov , a.v . , & dekel , a. 2002 , , 568 , 52 wechsler , r.h . ,
zentner , a.r . ,
bullock , j.s .
, kravtsov , a.v . , & allgood , b. 2006 , , 652 , 71 wetzel , a.r . , cohn , j.d . ,
white , m. , holz , d.e .
, & warren , m.s .
2007 , , 656 , 139 white , m. 2001 , , 367 , 27 yang , x. , mo , h.j .
, & van den bosch , f.c .
2006 , , 638 , l55 zhao , d.h .
, mo , h - j . , jing , y - p . , & boerner , g. 2003 , , 339 , 12 | galaxy cluster merger statistics are an important component in understanding the formation of large - scale structure .
cluster mergers are also potential sources of systematic error in the mass calibration of upcoming cluster surveys . unfortunately , it is difficult to study merger properties and evolution directly because the identification of cluster mergers in observations is problematic .
we use large n - body simulations to study the statistical properties of massive halo mergers , specifically investigating the utility of close halo pairs as proxies for mergers .
we examine the relationship between pairs and mergers for a wide range of merger timescales , halo masses , and redshifts ( @xmath0 ) .
we also quantify the utility of pairs in measuring merger bias .
while pairs at very small separations will reliably merge , these constitute a small fraction of the total merger population .
thus , pairs do not provide a reliable direct proxy to the total merger population .
we do find an intriguing universality in the relation between close pairs and mergers , which in principle could allow for an estimate of the statistical merger rate from the pair fraction within a scaled separation , but including the effects of redshift space distortions strongly degrades this relation .
we find similar behavior for galaxy - mass halos , making our results applicable to field galaxy mergers at high redshift .
we investigate how the halo merger rate can be statistically described by the halo mass function via the merger kernel ( coagulation ) , finding an interesting environmental dependence of merging : halos within the mass resolution of our simulations merge less efficiently in overdense environments .
specifically , halo pairs with separations less than a few @xmath1 are more likely to merge in underdense environments ; at larger separations , pairs are more likely to merge in overdense environments . |
charge - exchange between an atomic projectile and a surface plays a central role in surface science .
@xcite many surface diagnostics , for instance , secondary ion mass spectrometry @xcite or meta - stable atom de - excitation spectroscopy @xcite utilize surface - based charge - transfer processes . the same holds for plasma science .
surface - based production of negative hydrogen ions , for instance , is currently considered as a pre - stage process in neutral gas heating of fusion plasmas .
@xcite the operation modii of low - temperature gas discharges @xcite , which are main work horses in many surface modification and semiconductor industries , depend on secondary electron emission from the plasma walls and thus also on surface - based charge - transfer processes .
besides of their great technological importance , charge - transferring atom - surface collisions are however also of fundamental interest .
this type of collision couples a local quantum system with a finite number of discrete states the projectile to a large reservoir with a continuum of states the target .
irrespective of the coupling between the two , either due to tunneling or due to auger - type coulomb interaction , charge - transferring atom - surface collisions are thus perfect realizations of time - dependent quantum impurity systems .
@xcite by a judicious choice of the projectile - target combination as well as the collision parameters kondo - type features @xcite are thus expected as in any other quantum impurity system .
@xcite indeed a recent experiment by he and yarmoff @xcite provides strong evidence for electron correlations affecting the neutralization of positively charged strontium ions on gold surfaces .
the fingerprint of correlations could be the experimentally found negative temperature dependence of the neutralization probability .
it may arise @xcite from thermally excited conduction band holes occupying the strongly renormalized @xmath0 configuration of the projectile which effectively stabilizes the impinging ion and reduces thereby the neutralization probability .
the purpose of the present work is to analyze the he - yarmoff experiment @xcite from a genuine many - body theoretical point of view , following the seminal work of nordlander and coworkers @xcite as well as merino and marston @xcite and to provide theoretical support for the interpretation of the experiment in terms of a mixed - valence scenario .
we couch as usual the theoretical description of the charge - transferring atom - surface collision in a time - dependent anderson impurity model .
@xcite the parameters of the model are critical . to be as realistic as possible without performing an expensive ab - initio analysis of the ion - surface interaction we employ for the calculation of the model parameters gadzuk s semi - empirical approach @xcite based on image charges and hartree - fock wave functions for the projectile states .
@xcite the time - dependent anderson model , written in terms of pseudo - operators @xcite for the projectile states , is then subjected to a full quantum - kinetic analysis using contour - ordered green functions @xcite and a non - crossing approximation for the hybridization self - energies as originally proposed by nordlander and coworkers .
@xcite we apply the formalism to analyze , respectively , the neutralization of a strontium and a magnesium ion on a gold surface . for the mg : au system , which shows no evidence for mixed - valence correlations affecting the charge - transfer between the surface and the projectile
, we find excellent agreement between theory and experiment .
for the sr : au system , in contrast , we could reproduce only the correct order of magnitude of the neutralization probability .
its temperature dependence could not be reproduced .
our modeling shows however that a mixed - valence scenario could in principle be at work .
for the material parameters best suited for the description of the sr : au system they are however not strong enough to affect the neutralization probability also quantitatively .
the outline of our presentation is as follows .
in the next section we describe the time - dependent anderson model explaining in particular how we obtained the parameters characterizing it .
section [ quantumkinetics ] concerns the quantum kinetics and presents the set of coupled two - time integro - differential equations which have to be solved for determining the probabilities with which the various charge states of the projectile occur .
they form the basis for the analysis of the temperature dependence of the neutralization probability .
numerical results for a strontium as well as a magnesium ion hitting a gold surface are presented , discussed , and compared to experimental data in sect .
[ results ] .
concluding remarks are given in sect .
[ conclusions ] .
when an atomic projectile approaches a surface its energy levels shift and broaden due to direct and exchange coulomb interactions with the surface .
since the target and the projectile are composite objects the calculation of these shifts and broadenings from first principles is a complicated problem .
@xcite we follow therefore gadzuk s semi - empirical approach .
@xcite from our previous work on secondary electron emission due to de - excitation of meta - stable nitrogen molecules on metal @xcite and dielectric @xcite surfaces we expect the approach to give reasonable estimates for the level widths as well as the level positions for distances from the surface larger than a few bohr radii .
in addition , the approach has a clear physical picture behind it and is thus intuitively very appealing .
the essence of the model is illustrated in fig .
it shows for the particular case of a strontium ion hitting a gold surface the energy levels of the projectile closest to the fermi energy of the target .
quite generally , for alkaline - earth ( ae ) ions the first and the second ionization levels are most important . identifying the positive ion ( @xmath1 ) with a singly occupied impurity and the neutral atom ( @xmath2 ) with a doubly occupied impurity
, the projectile can be modelled as a non - degenerate , asymmetric anderson model with on - site energies @xmath3 where @xmath4 and @xmath5 are , respectively , the first and second ionization energy far away from the surface while @xmath6 is the distance of the metal s image plane from its crystallographic ending at @xmath7 .
the on - site coulomb repulsion @xmath8 would be the difference of the two energies .
table [ parameters ] summarizes the material parameters required for the modeling of the neutralization of strontium and magnesium ions on a gold surface .
.material parameters for magnesium , strontium and gold : @xmath9 and @xmath10 are the first and the second ionization energy , @xmath11 and @xmath12 are the effective charges to be used in the calculation of the hybridization matrix element ( viz : eq . ) , @xmath13 is the work function , @xmath14 the fermi energy , @xmath6 the position of the image plane in front of the surface , and @xmath15 is the effective mass of an electron . [ cols="^,^,^,^,^,^,^,^,^",options="header " , ] the @xmath16dependent shifts of the ionization levels can be obtained as the energy gain of a virtual process moving the configuration under consideration from the actual position @xmath17 to @xmath18 , reducing its electron occupancy by one , and then moving it back to position @xmath17 , taking into account in both moves if present
image interactions due to the charge state of the final and initial configurations with the metal .
@xcite for the upper level , @xmath19 , that is , the @xmath20 configuration the cycle is @xmath21 , whereas for the lower level , @xmath22 , that is , the @xmath0 configuration the cycle is @xmath23 .
ion on a gold surface .
the two ionization energies , @xmath24 and @xmath25 , standing for the projectiles @xmath20 and @xmath0 configuration , respectively , shift due to the image interaction with the surface . far away from the surface
the two energies merge , respectively , with the first @xmath26 and the second ( @xmath10 ) ionization energy of a strontium atom .
the image interaction also leads to a hybridization of the sr states with the conduction band states of the surface which is characterized by a step potential at @xmath7 whose depth is the sum of the work function @xmath27 and the fermi energy @xmath28 . for simplicity
the broadening is not shown .
indicated however is the trajectory @xmath29 of the ion .
important points along the trajectory are @xmath30 , the turning point , and @xmath31 , the point where the first ionization level crosses the fermi energy . ] to set up the hamiltonian we also need the wave functions for the projectile states . for the upper level we use the ( ns ) hartree - fock wave function of an @xmath32 atom while for the lower level we use the ( ns ) hartree - fock wave function of an @xmath1 ion . according to clementi and
roetti @xcite both can be written in the form @xmath33 with @xmath34 , and @xmath35 tabulated parameters and @xmath36 the spherical harmonics with @xmath37 . ) for the mg : au ( solid lines ) and the sr : au ( dashed lines ) system . ] for simplicity we assume the projectile to approach the surface from @xmath18 on a perpendicular trajectory , @xmath38 with the turning point @xmath30 reached at time @xmath39 and @xmath40 the velocity of the projectile .
the lateral motion of the projectile is thus ignored . to be consistent with this simple trajectory we also neglect the lateral variation of the potential characterizing the metal surface .
the electrons of the metal are thus simply described in terms of a potential step at @xmath7 with depth @xmath41 , where @xmath27 is the work function of the metal and @xmath28 is its fermi energy measured from the bottom of the conduction band ( see table [ parameters ] ) , leading to @xmath42 \theta(-z ) \bigr\}~,\end{aligned}\ ] ] for the energies and wave functions of the conduction band electrons ; @xmath43 is the spatial width of the step ( drops out in the final expressions ) and @xmath44 with @xmath45 are the reflection and transmission coefficients of the potential step .
while the projectile is on its trajectory its ionization levels hybridize with the conduction band . the matrix element for this process
is given by @xcite @xmath46 where the potential between the two wave functions is the residual coulomb interaction of the valence electron with the core of the projectile located at @xmath47 .
the matrix element can be transformed to a level width @xmath48 which is an important quantity . the charge @xmath49 in eq .
is the charge of the nucleus screened by all the electrons of the projectile except of the valence electron under consideration .
for the hybridization of the lower level , the second ionization level , @xmath50 while for the hybridization of the upper level , the first ionization level , @xmath51 , where @xmath52 is slater s shielding constant due to the second electron in the @xmath53-valence shell .
@xcite in fig .
[ levelwidths ] we show the level widths calculated form eq . with @xmath54 set , respectively , to @xmath24 and @xmath25 , for magnesium and strontium using the parameters of table [ parameters ] .
most probably we overestimate the widths close to the surface . to what extent , however , only precise calculations of the kind performed for alkaline ions by nordlander and tully can show . @xcite , @xmath55 , and @xmath56 are , respectively , the ( pseudo ) probabilities with which the @xmath1 , the @xmath2 , and the @xmath57 configuration occur . ] using coleman s pseudo - particle representation @xcite for the projectile configurations illustrated in fig .
[ pseudostates ] , the hamiltonian describing the interaction of an ae projectile with a metal surface can be written as @xcite @xmath58d^\dagger d^{{\phantom{\dag}}}\nonumber\\ & + \sum_{\vec{k}\sigma}\varepsilon_{\vec{k } } c_{\vec{k}\sigma}^\dagger c_{\vec{k}\sigma}^{{\phantom{\dag}}}+\sum_{\vec{k}\sigma}\big[v_{\vec{k}}(t)c_{\vec{k}\sigma}^\dagger e^\dagger p_\sigma^{{\phantom{\dag}}}+ h.c .
\big]\nonumber\\ & + \sum_{\vec{k}\sigma}\big[v_{\vec{k}}(t)c_{\vec{k}\sigma}^\dagger d^{{\phantom{\dag}}}p_{-\sigma}^\dagger + h.c .
\big ] \label{hamiltonian}\end{aligned}\ ] ] with @xmath59 , @xmath60 , and @xmath61 denoting , respectively , the creation operators for an empty ( @xmath57 ) , a doubly occupied ( @xmath62 ) , and a singly occupied ( @xmath63 ) projectile . since the projectile can be only in either one of these configurations , the hamiltonian has to be constrained by @xcite @xmath64 this completes the description of the model .
combined with measured projectile velocities the model describes the charge - transfer responsible for the neutralization of alkaline - earth ions on noble metal surfaces . .
diagrams ( a ) and ( b ) give , respectively , the self - energies @xmath65 and @xmath66 for the green function @xmath67 . the self - energies @xmath68 and @xmath69 for the green functions @xmath70 and @xmath71 , respectively , are shown in ( c ) and ( d ) . ]
to calculate the neutralization probability for the ae ion hitting the metal surface we follow nordlander and coworkers @xcite and set up quantum - kinetic equations for contour - ordered green functions @xcite describing the empty , singly , and doubly occupied projectile .
we denote these functions , respectively , by @xmath72 , @xmath73 , and @xmath74 and write their analytic pieces in the form @xmath75 \bar{h}^{\rm r}(t , t^\prime)~,\\
h^{\gtrless}(t , t^\prime ) & = \exp[-i\!\int_{t^\prime}^t \!\!d\bar{t}\varepsilon(\bar{t } ) ] \bar{h}^{\gtrless}(t , t^\prime)~,\end{aligned}\ ] ] where @xmath76 can be any of the three green functions and @xmath54 is , depending on the green function , either identically @xmath77 , @xmath25 , or @xmath78 . using this notation and calculating the self - energies @xmath68 , @xmath65 , @xmath66 , and @xmath69 in the non - crossing approximation diagrammatically shown in fig .
[ nca ] leads after application of the langreth - wilkins rules @xcite and the projection to the @xmath79 subspace @xcite to @xcite @xmath80 and @xmath81^ * -\sum_\sigma\int^{t}_{-\infty } \!\!d\bar{t } \bar{k}_{\varepsilon_0}^{<}(\bar{t},t)\bar{p}_\sigma^{\rm r}(t,\bar{t})\bar{e}^{<}(\bar{t},t^\prime)~ , \label{elt}\\ \frac{\partial}{\partial t } \bar{p}_\sigma^{<}(t , t^\prime ) & = \int^{t^\prime}_{-\infty } \!\!d\bar{t } \bar{k}_{\varepsilon_0}^{<}(t,\bar{t})\bar{e}^{<}(t,\bar{t})[\bar{p}_\sigma^{\rm r}(t^\prime,\bar{t})]^ * + \int^{t^\prime}_{-\infty } \!\!d\bar{t } \bar{k}_{\varepsilon_u}^{>}(\bar{t},t)\bar{d}^{<}(t,\bar{t})[\bar{p}_\sigma^{\rm r}(t^\prime,\bar{t})]^ * \nonumber\\ & -\int^{t}_{-\infty } \!\!d\bar{t } \bar{k}_{\varepsilon_0}^{>}(t,\bar{t})\bar{e}^{\rm r}(t,\bar{t})\bar{p}_\sigma^{<}(\bar{t},t^\prime ) -\int^{t}_{-\infty } \!\!d\bar{t } \bar{k}_{\varepsilon_u}^{<}(\bar{t},t)\bar{d}^{\rm r}(t,\bar{t})\bar{p}_\sigma^{<}(\bar{t},t^\prime)~ , \label{plt}\\ \frac{\partial}{\partial t } \bar{d}^{<}(t , t^\prime ) & = \sum_\sigma\int^{t^\prime}_{-\infty } \!\!d\bar{t } \bar{k}_{\varepsilon_u}^{<}(t,\bar{t})\bar{p}_{-\sigma}^{<}(t,\bar{t})[\bar{d}^{\rm r}(t^\prime,\bar{t})]^ * -\sum_\sigma\int^{t}_{-\infty } \!\!d\bar{t } \bar{k}_{\varepsilon_u}^{>}(t,\bar{t})\bar{p}_{-\sigma}^{\rm r}(t,\bar{t})\bar{d}^{<}(\bar{t},t^\prime)~ \label{dlt}\end{aligned}\ ] ] with @xmath82 and @xmath83 f^{\,\gtrless}(t - t^\prime)~,\end{aligned}\ ] ] where @xmath84 is the fourier transform of the fermi function @xmath85 defined by @xmath86~. \end{aligned}\ ] ] the function @xmath87 , which contains the temperature dependence , entails an approximate momentum summation . from the diagrams shown in fig .
[ nca ] one initially obtains @xmath88\end{aligned}\ ] ] with an energy integration extending over the range of the conduction band and @xmath89 given by eq . with @xmath54
replaced by the integration variable @xmath90 . to avoid the numerically costly energy integration nordlander and coworkers employed two different approximations : in ref .
they replaced @xmath89 by an average over the energy range of the conduction band while in ref .
they replaced it by @xmath91 with @xmath54 set to @xmath25 or @xmath24 depending on which state is considered in the hybridization self - energy . using the latter leads to @xmath92 and eventually to @xmath93 as given in eq . .
the subscript @xmath90 indicates now not an integration variable but the functional dependence on @xmath54 .
we employ this form but keep in mind that it is an approximation to the non - crossing self - energies . the instantaneous ( pseudo )
occurrence probabilities for the projectile configurations @xmath57 , @xmath63 , and @xmath2
are then given by @xmath94 respectively , where we refer to all of them as ( pseudo ) occurrence probabilities also strictly speaking @xmath55 and @xmath56 are true ones and only @xmath95 is a pseudo occurrence probability in the sense that the true probability with which the @xmath1 configuration occurs is @xmath96 .
@xcite sometimes we will also refer to @xmath56 , @xmath95 , and @xmath55 simply as ( pseudo ) occupancies . for the ae ion
the probability for neutralization at the surface ( wall recombination ) is the probability for double occupancy after the completion of the trajectory , that is , @xmath97 subject to the initial conditions @xmath98 and @xmath99 .
we solve the two coupled sets of integro - differential equations and on a two - dimensional time grid setting @xmath100 for the retarded green functions and @xmath101 for the less - than green functions using basically the same numerical strategy as shao and coworkers .
@xcite plane over which eqs.([er])([dr ] ) and ( [ elt])([dlt ] ) have to be integrated subject to the boundary conditions ( [ bcr ] ) and ( [ bcelt])([bcdlt ] ) , respectively , in order to determine the retarded and less - than green functions at @xmath102 .
the triangular ( rectangular ) region denotes the domain required for the calculation of the retarded ( less - than ) green functions . ]
due to the intertwining of the time integrations the integration domains for the retarded green functions are triangular whereas for the less - than green function they are rectangular as shown in fig .
[ intdomain ] .
the size of the time - grid as well as the discretization depend on the velocity of the projectile and the maximum distance it has from the surface .
for the he - yarmoff experiment the velocities are on the order of @xmath103 in atomic units .
the maximum distance from which the ion starts its journey can be taken to be 20 bohr radii . at this distance
the coupling between the surface and the ion is vanishingly small .
we empirically found the algorithm to converge for a @xmath104 grid with @xmath105 . since the green functions are complex the computations are time and memory consuming .
we now analyze the he - yarmoff experiment @xcite quantitatively from a many - body theoretical point of view . for that purpose
we combine the model developed in sect . [ model ] with the quantum - kinetics described in sect .
[ quantumkinetics ] .
besides the parameters given in table [ parameters ] we also need the velocity of the projectile . in general
, the velocity will be different on the in- and outgoing branch of the trajectory .
the outgoing branch , however , determines the final charge state of the projectile .
we take therefore for both branches the normal component of the experimentally measured post - collision velocity .
if not noted otherwise all quantities are in atomic units , that is , energies are measured in hartrees and lengths in bohr radii .
k as a function of time .
the projectile starts at @xmath106 and @xmath107 with velocity @xmath108 , reaches at @xmath39 the turning point @xmath109 , and approaches at @xmath110 again @xmath111 .
the ionization levels ( solid lines ) are broadened according to @xmath112 ( dashed lines ) with @xmath113 as shown in fig .
[ levelwidths ] .
lower panel : instantaneous ( pseudo ) occurrence probabilities along the trajectory for the @xmath114 , the @xmath115 , and the @xmath116 configurations . initially , at time @xmath106
, the projectile is in the @xmath114 configuration .
the neutralization probability in this particular case is @xmath117 ( solid bullet ) . ]
first , we discuss the mg : au system . in fig .
[ occurrencemg ] we show the time - dependence of the broadened ionization levels , @xmath19 and @xmath22 , together with the instantaneous ( pseudo ) occurrence probabilities @xmath118 , and @xmath56 for the @xmath114 , the @xmath119 , and the @xmath116 configuration , respectively . negative and positive times denote the in- and outcoming branch of the trajectory .
the velocity @xmath108 and the surface temperature @xmath120 .
initially , the projectile is in the @xmath114 configuration , that is , the lower level @xmath22 , representing single occupancy , is occupied while the upper level @xmath19 , representing double occupancy , and thus the @xmath119 configuration , is empty . while the projectile is on its way through the trajectory the ionization levels shift and broaden . as a result the occupancies change .
the neutralization probability is then the probability for double occupancy at the end of the trajectory . for the particular case of the mg : au system the first ionization level , @xmath19 , that is
, the level which has to accept an electron in order to neutralize the ion , is below the fermi energy of the metal throughout the whole trajectory .
the broadening is also rather weak .
it only leaks for a very short time span above the fermi energy . as a result
, the magnesium ion can efficiently soak in a second electron while the electron already present due to the initial condition is basically frozen in the second ionization level .
the electron captured from the metal has moreover a strong tendency to stay on the projectile .
it only has a chance to leave it in the short time span where the instantaneous broadening @xmath121 is larger than @xmath122 . the neutralization probability is thus expected to be close to unity .
indeed , we find for the situation shown in fig .
[ occurrencemg ] @xmath123 ( solid bullet in fig .
[ occurrencemg ] ) . the temperature dependence of @xmath124 is shown in fig .
[ alphamg ] . in accordance with experiment
we find @xmath124 essentially to be independent of temperature .
this is expected because both ionization levels , @xmath19 and @xmath22 , are below the fermi energy and their broadening is too small to allow a charge - transfer from the projectile to empty conduction band states of the surface .
notice , the excellent agreement between theory and experiment indicating that the semi - empirical model we developed in sect .
[ model ] captures the essential features of the charge - transfer pretty well . for a @xmath114 ion hitting with @xmath108 a gold surface .
the turning point @xmath125 .
also shown are experimental data from ref . .
] after the successful description of the mg : au system let us now turn to the sr : au system . in fig . [ occurrencesr ]
we again plot as a function of time the broadened ionization levels and the ( pseudo ) occurrence probabilities for the three configurations of the projectile . as it was the case for mg : au , the configuration of the projectile , which initially was in the configuration representing single occupancy , changes along the trajectory .
the changes are however more subtle .
the reason is the level structure .
in contrast to the mg : au system , the ionization levels are now closer to the fermi energy of the surface .
the first ionization level @xmath19 even crosses the fermi energy with far reaching consequences . the part of the trajectory where @xmath19 is below the fermi energy , that is , the region where the neutral atom would be energetically favored , the broadening is very small , indicating negligible charge - transfer from the metal to the ion and hence a stabilization of the ion due to lack of coupling . when the broadening and thus the coupling is large @xmath19 is above the fermi level . in this part of the trajectory
the ion is energetically stabilized .
the first ionization level of strontium can capture an electron from the metal only in the time span where @xmath126 .
the neutralization probability of a strontium ion should be thus much smaller than the one for a magnesium ion .
indeed we find @xmath127 which is much smaller than unity ( solid bullet in fig .
[ occurrencesr ] ) .
k as a function of time .
the projectile starts at @xmath128 and @xmath107 with velocity @xmath129 , reaches at @xmath39 the turning point @xmath109 , and approaches at @xmath130 again @xmath131 . the levels ( solid lines )
are broadened according to @xmath112 ( dashed lines ) with @xmath113 as shown in fig .
[ levelwidths ] .
lower panel : instantaneous ( pseudo ) occurrence probabilities along the trajectory for the @xmath132 , the @xmath133 , and the @xmath134 configurations . initially , at time @xmath128 , the projectile is in the @xmath132 configuration .
the neutralization probability in this particular case is @xmath135 ( solid bullet ) . ] due to the shift and broadening of the first ionization level @xmath19 it is clear that a strontium ion can not as efficiently neutralize on a gold surface as a magnesium ion .
this sets the scale of @xmath124 .
in addition , and in great contrast to magnesium , the second ionization level @xmath22 is however also close to the fermi energy . in those parts of the trajectory for which @xmath136 it can affect the charge - transfer between the metal and the projectile .
in fact , taken by itself , it should stabilize the ion and hence decrease the neutralization probability .
@xcite qualitatively , this can be understood from a density of states argument . from the upper panel of fig .
[ occurrencesr ] we can infer that the broadened second ionization level is cut by the fermi energy in the upper half of its local density of states .
hence , close to the surface holes start to occupy the second ionization level at energies where the local density of states is higher than at the energies where electrons are transferred .
increasing temperature enhances thus the tendency of electron loss from the second ionization level . without interference from the first ionization level the neutralization probability should thus go down with temperature . for a @xmath132 ion hitting with @xmath129 a gold surface .
the turning point @xmath125 . also shown are experimental data from ref . .
] that the second ionization level of sr comes close to the fermi energy of au most probably led he and yarmoff @xcite to suggest that the neutralization of strontium ions on gold surfaces is dominated by electron correlations .
indeed the experimentally found negative temperature dependence of @xmath124 above @xmath137 seems to support their conclusion .
however , the temperature dependence of @xmath124 we obtain and which we plot in fig .
[ alphasr ] , does not show this behavior , at least , for the material parameters of table [ parameters ] and the experimentally measured post - collision velocity .
the reason for the discrepancy between the measured and the calculated data is unclear . the material parameters seem to be reasonable since the theoretical results have the correct order of magnitude .
it could be however that the temperature - induced transfer of holes to the second ionization level is overcompensated by the electron - transfer to the first ionization level . in the region where charge - transfer is strongest the two ionization levels overlap .
the absence of energy separation together with the conditional temporal weighting due to the dynamics of the collision process makes it very hard to tell a priori which process will win and manifest itself in the measured neutralization probability .
so far , the discussion of the data left out the possibility of a correlation - induced sharp resonance in the vicinity of the fermi energy , that is , the key feature of kondo - type physics .
the numerical results seem to suggest that either there is no resonance or it does not affect the neutralization process .
however , from the data itself we can not determine which one is the case .
we can thus not decide whether the sr : au system is in a correlated regime or not and hence whether an interpretation of the experimental data in terms of a mixed - valence scenario is in principle plausible or has to be dismissed .
a rigorous way to decide this would be to calculate the instantaneous spectral functions for the projectile and to look for sharp resonances in the vicinity of the fermi energy .
this is beyond the scope of the present work . to get at least a qualitative idea about in what regime
the strontium projectile might be along its trajectory we plot in fig .
[ estarsr ] , following merino and marston , @xcite haldane s scaling invariant , @xcite @xmath138 as a function of time travelled along the outgoing branch of the trajectory . in the perturbative regime which is strictly applicable only far away from the surface @xmath139 can be interpreted as the renormalized second ionization level and @xmath140 .
@xcite for @xmath141 the projectile is likely to be in the mixed - valence regime .
@xcite since @xmath139 comes very close to the fermi energy holes are expected to transfer in the mixed - valence regime very efficiently to the projectile . in situations where the projectile stays sufficiently long in the mixed - valence regime before @xmath139 crosses the fermi energy double occupancy and hence the neutralization probability should be suppressed with increasing temperature . as can be seen in fig .
[ estarsr ] close enough to the surface the strontium projectile is indeed in the mixed - valence regime .
for the material parameters given in table [ parameters ] and the experimental value for the projectile velocity the time - span however is rather short .
most probably this is the reason why we do not see any reduction of @xmath124 with temperature for the parameters we think to be best suited for the sr : au system . since the experimental data are unambiguous , this indicates perhaps the need for a precise first - principle calculation of the model parameters .
alternative interpretations of the experimental results can however not be ruled out . ) and bare ( @xmath22 ) second ionization level measured from the fermi energy as a function of time traveled along the outgoing branch of the trajectory for the @xmath142:au system with @xmath129 and @xmath120 .
also shown as a function of time is the scaling invariant @xmath143 . in the region
for which @xmath141 the system is likely to be in the mixed - valence region .
the material parameters are as given in table [ parameters ] and @xmath129 . ]
motivated by claims that the neutralization of strontium ions on gold surfaces is affected by electron correlations we set up a semi - empirical model for charge - transferring collisions between alkaline - earth projectiles and noble metal surfaces . the surface is simply modelled by a step potential while the projectile is modelled by its two highest ionization levels which couple to the surface via gadzuk s image - potential - based projectile - surface interaction . to calculate the neutralization probability we employed a pseudo - particle representation of the projectile s charge states and quantum - kinetic equations for the retarded and less - than green functions of the projectile as initially suggested by nordlander , shao and langreth . besides the non - crossing approximation for the self - energies and an approximate momentum summation
no further approximations are made .
the quantum - kinetic equations are numerically solved on a two - dimensional time - grid using essentially the same strategy as shao and coworkers .
the absolute values for the neutralization probability we obtain are in good agreement with experimental data , especially for the mg : au system , but also for the sr : au system , although for the latter we could not reproduce the temperature dependence of the neutralization probability .
our calculations can thus not decide whether the he - yarmoff experiment can be interpreted in terms of a mixed - valence scenario . from the instantaneous values of haldane s scaling invariant we see
however that the sr : au system could be in the mixed - valence regime .
the mechanism for a negative temperature dependence , that is , the possibility of efficiently transferring holes to the second ionization level , is thus in principle present . for the material parameters however most appropriate for sr : au the negative temperature dependence arising from this channel seems to be overcompensated by the positive temperature dependence of the electron - transfer to the first ionization level . to proof that he and yarmoff have indeed seen for the first time
mixed - valence correlations affecting charge - transfer between an ion and a surface requires therefore further theoretical work .
m. p. was funded by the federal state of mecklenburg - western pomerania through a postgraduate scholarship within the international helmholtz graduate school for plasma physics .
in addition , support from the deutsche forschungsgemeinschaft through project b10 of the transregional collaborative research center sfb / trr24 is greatly acknowledged . | motivated by experimental evidence for a mixed - valence state to occur in the neutralization of strontium ions on gold surfaces we analyze this type of charge - transferring atom - surface collision from a many - body theoretical point of view using quantum - kinetic equations together with a pseudo - particle representation for the electronic configurations of the atomic projectile .
particular attention is paid to the temperature dependence of the neutralization probability which experimentally seems to signal mixed - valence - type correlations affecting the charge - transfer between the gold surface and the strontium projectile .
we also investigate the neutralization of magnesium ions on a gold surface which shows no evidence for a mixed - valence state .
whereas for magnesium excellent agreement between theory and experiment could be obtained , for strontium we could not reproduce the experimental data .
our results indicate mixed - valence correlations to be in principle present , but for the model mimicking most closely the experimental situation they are not strong enough to affect the neutralization process quantitatively . |
let @xmath0 be a proper birational map between smooth varieties over a field of characteristic zero .
a commutative diagram @xmath1 where @xmath2 and @xmath3 are sequences of blowups of smooth centers , is called a strong factorization of @xmath4 .
the existence of a strong factorization is an open problem in dimension @xmath5 and higher .
the local version of the strong factorization conjecture replaces the varieties by local rings dominated by a valuation on their common fraction field , and the smooth blowups by monoidal transforms along the valuation .
the local strong factorization was proved by c. christensen @xcite in dimension @xmath6 for certain valuations .
a complete proof of the @xmath6-dimensional case was given by s. d. cutkosky in @xcite , where he also made considerable progress towards proving the conjecture in general .
we prove the local factorization conjecture in any dimension ( see section [ sec - ring ] for notation ) : [ thm1 ] let @xmath7 and @xmath8 be excellent regular local rings containing a field @xmath9 of characteristic zero .
assume that @xmath7 and @xmath8 have a common fraction field @xmath10 and @xmath11 is a valuation on @xmath10 .
then there exists a local ring @xmath12 , obtained from both @xmath7 and @xmath8 by sequences of monoidal transforms along @xmath11 .
the toric version of the strong factorization problem considers two nonsingular fans @xmath13 and @xmath14 with the same support and asks whether there exists a common refinement @xmath15 @xmath16 obtained from both @xmath13 and @xmath14 by sequences of smooth star subdivisions .
again , this is not known in dimension @xmath6 or higher .
the local toric version replaces a fan by a single cone and a smooth star subdivision of the fan by a smooth star subdivision of the cone together with a choice of one cone in the subdivision .
we assume that the choice is given by a vector @xmath17 in the cone : we choose a cone in the subdivision containing @xmath17 .
if @xmath17 has rationally independent coordinates , then it determines a unique cone in every subdivision ( all cones are rational ) .
we call such a vector @xmath17 a _ valuation _ and the subdivision with a choice of a cone a _ subdivision _ along the valuation .
we prove : [ thm2 ] let @xmath18 and @xmath19 be nonsingular cones , and let @xmath20 be a vector with rationally independent coordinates .
then there exists a nonsingular cone @xmath21 obtained from both @xmath18 and @xmath19 by sequences of smooth star subdivisions along @xmath17 .
the proof of theorem [ thm2 ] is a generalization of the proof given by c. christensen @xcite in dimension @xmath6 .
theorem [ thm1 ] follows directly from theorem [ thm2 ] and the monomialization theorem proved by s. d. cutkosky @xcite .
theorem [ thm1 ] ia also stated in @xcite , but the proof refers to the strong factorization theorem in @xcite which contains a gap @xcite .
we use the same reduction to the toric case , but replace the reference to strong factorization by theorem [ thm2 ] .
one can define a more general version of local toric factorization .
consider a game between two players @xmath22 and @xmath23 , where the player @xmath22 subdivides the cone @xmath19 or @xmath18 and the player @xmath23 chooses one cone in the subdivision ( and renames it again @xmath19 or @xmath18 ) .
then the strong factorization conjecture states that @xmath22 always has a winning strategy : after a finite number of steps either @xmath24 or the interiors of @xmath19 and @xmath18 do not intersect . the proof of theorem [ thm2 ] given in section [ sec - toric ] does not extend to this more general case . a positive answer to the global strong factorization conjecture for toric varieties would imply that @xmath22 always has a winning strategy .
conversely , a counterexample to the local factorization problem would give a counterexample to the global strong factorization conjecture .
* acknowledgments .
* i have benefited a great deal from discussions of the factorization problem with dan abramovich , kenji matsuki and jarosaw wodarczyk .
it was jarosaw s suggestion to look for a counterexample in dimension @xmath25 that motivated the current proof .
let @xmath26 be a lattice and @xmath18 a rational polyhedral cone in @xmath27 generated by a finite set of vectors @xmath28 @xmath29 we say that @xmath18 is _ nonsingular _ if it can be generated by a part of a basis of @xmath30 .
a nonsingular @xmath31-dimensional cone has a unique set of minimal generators @xmath32 , and we write @xmath33 we consider nonsingular cones only .
when we draw a picture of a cone , we only show a cross - section .
thus , a @xmath6-dimensional cone is drawn as a triangle .
let @xmath34 be a nonsingular @xmath35-dimensional cone , and let @xmath36 be a vector @xmath37 such that @xmath38 are linearly independent over @xmath39 . if @xmath40 , then precisely one of the cones @xmath41 contains @xmath17 .
the cone containing @xmath17 is called a _ star subdivision of @xmath18 at @xmath42 along @xmath17_.
the subdivision is again a nonsingular cone .
we often denote a star subdivision of a cone @xmath18 again @xmath18 , and name its generators also @xmath43 .
let us consider the situation of theorem [ thm2 ] .
it is easy to see that after star subdividing @xmath19 sufficiently many times we may assume that @xmath44 .
we say that a configuration @xmath44 is _ factorizable _ if the statement of theorem [ thm2 ] holds .
we say that @xmath44 is _ directly factorizable _ if the statement of theorem [ thm2 ] holds with @xmath45 .
the vector @xmath17 is not needed for direct factorizability .
the following lemma is well - known : [ lem - dim2 ] if the dimension @xmath46 , then @xmath44 is directly factorizable .
[ lem - dim3 ] let @xmath5 and consider @xmath47 , where @xmath48 are nonsingular cones such that @xmath49 are linearly dependent .
then @xmath50 is directly factorizable
. * proof .
* let @xmath51 be the quotient map .
we claim that @xmath52 are nonsingular cones with respect to the lattice @xmath53 .
this is clear for the cone @xmath18 ; for @xmath19 note that the generators @xmath54 of @xmath30 map to generators of @xmath53 .
( more precisely , @xmath55 , @xmath56 , where @xmath57 is primitive , @xmath58 . )
now we apply lemma [ lem - dim2 ] to the configuration @xmath59 .
then after a finite sequence of star subdivisions of @xmath18 at vectors lying in @xmath60 , we may assume that @xmath61 if we express @xmath62 , then it follows from the nonsingularity of @xmath19 that @xmath63 .
in other words , the cone @xmath19 lies in the subdivision @xmath64 of @xmath18 . performing a sequence of such star subdivisions , we get to the situation where @xmath65 .
finally , @xmath66 is strongly factorizable by lemma [ lem - dim2 ] , thus a sequence of star subdivisions of @xmath18 at vectors lying in @xmath67 finishes the proof . by the previous lemma , to show that @xmath50 is factorizable , we have to find a sequence of star subdivisions of @xmath19 such that the condition of the lemma is satisfied . we prove this in any dimension .
[ lem - align ] let @xmath68 and consider a configuration @xmath69
. there exists a cone @xmath70 , obtained from @xmath19 by a sequence of smooth star subdivisions along @xmath17 , such that @xmath49 are linearly dependent .
moreover , one can find @xmath71 such that @xmath49 satisfy the relation @xmath72 * proof .
* the first part of the proof is again due to c. christensen .
let us start with the case @xmath5 and prove the first half of the lemma .
the algorithm for constructing @xmath71 is as follows .
let @xmath51 be the projection and let the generators @xmath54 of @xmath19 be ordered so that @xmath73 .
if @xmath74 or @xmath75 , then we are done .
otherwise star subdivide @xmath19 at @xmath76 and repeat . to see that this algorithm always terminates , let @xmath77 be defined by : @xmath78 here @xmath79 , @xmath80 , and @xmath81 , @xmath82 .
then the algorithm can be described as follows .
consider the matrix @xmath83 if some @xmath84 , then we are done .
otherwise , choose columns @xmath85 and @xmath86 such that @xmath87 and @xmath88 have the same sign and subtract the @xmath85th column from the @xmath86th if @xmath89 and @xmath86th column from the @xmath85th if @xmath90 .
since we always choose columns where @xmath87 and @xmath88 have the same sign , it is clear that @xmath91 does not increase in this process , and it suffices to prove that either the algorithm terminates or @xmath91 drops after a finite number of steps .
suppose that @xmath92 , and @xmath93 does not change as we run the algorithm .
then @xmath94 also does not change , and every time we choose columns @xmath85 and @xmath6 , we subtract @xmath94 from @xmath95 .
it is clear that columns @xmath96 and @xmath97 can be chosen only a finite number of times in a row , hence we choose column @xmath6 infinitely many times .
since we can not subtract @xmath94 from @xmath98 or @xmath99 infinitely many times and have a positive result , we get a contradiction .
this proves the first half of the lemma for @xmath5 .
next let us prove the `` moreover '' part for @xmath5 .
we start with a matrix @xmath100 if also @xmath101 , then by nonsingularity of @xmath19 we have @xmath102 .
we choose columns @xmath96 and @xmath97 the necessary number of times to get @xmath103 : @xmath104 if both @xmath105 and @xmath106 are nonzero then they must have different signs . hence ,
if @xmath107 then we are done . otherwise , since @xmath108 , we may assume that @xmath109 .
we perform star subdivisions of @xmath19 by choosing columns @xmath97 and @xmath6 the necessary number of times to get to the matrix @xmath110 after this , we run the algorithm as before . for instance , since @xmath105 and @xmath111 have the same sign , at the next step we choose columns @xmath96 and @xmath6 . if we subtract the third column from the first , then latexmath:[$\max_i first column from the third .
as before , if @xmath91 does not decrease , then we are subtracting @xmath98 from @xmath99 or @xmath94 infinitely many times , and this gives a contradiction .
for @xmath113 we have a matrix @xmath114 we can apply the @xmath5 case to the last three columns and achieve @xmath115 ; then apply the same algorithm to columns @xmath116 to get @xmath117 , and so on , until all but two of the @xmath87 are nonzero . to prove the second half of the lemma
, we apply the @xmath5 case to three columns , including the ones with @xmath118 .
* proof of theorem [ thm2 ] .
* we may assume that @xmath119 , and using lemma [ lem - align ] , we may also assume that @xmath49 satisfy the relation @xmath120 let @xmath51 be the projection
. then @xmath121 are both nonsingular with respect to the lattice @xmath53 .
the relation @xmath122 implies that @xmath123 is a minimal generator of @xmath124 .
in particular , @xmath125 restricts to isomorphisms of cones and lattices : @xmath126 by induction on the dimension @xmath35 , we have a factorization of @xmath127 . unlike the case @xmath5
, we may also have to subdivide @xmath124 .
consider a star subdivision of @xmath124 at @xmath128 , @xmath129 , and define @xmath130 and @xmath131 by : @xmath132 now star subdividing @xmath19 first at @xmath130 and then at @xmath131 along @xmath17 , the resulting cone again satisfies the relation @xmath122 ( after possibly reordering the generators ) , and its image under @xmath125 is the star subdivision of @xmath124 at @xmath133 along @xmath134 . in other words ,
every star subdivision of @xmath124 can be lifted to a subdivision of @xmath19 .
thus after a finite sequence of star subdivisions of @xmath19 we may assume that @xmath135 is directly factorizable .
the remaining proof is the same as in the @xmath6-dimensional case .
star subdividing @xmath18 at vectors lying in the face @xmath136 , we may assume that @xmath137 if @xmath138 , then @xmath139 for @xmath140 , hence after star subdividing @xmath18 at vectors lying in the face @xmath141 , we may assume that @xmath142 for @xmath143 .
now @xmath144 are nonsingular cones , hence directly factorizable by lemma [ lem - dim2 ] .
a sequence of star subdivisions of @xmath18 at vectors lying in the face @xmath145 finishes the proof .
j. wodarczyk has noted that is makes sense to consider the local toric factorization problem also for a vector @xmath17 with rationally dependent coordinates , and this problem can be reduced to the rationally independent case .
we bring here an argument for such a reduction .
similar reduction appears in s. d. cutkosky s proof of the monomialization theorem @xcite . consider a nonsingular @xmath35-dimensional cone @xmath18 and a vector @xmath36 , with possibly rationally dependent coordinates .
a star subdivision of @xmath18 along @xmath17 is a star subdivision of @xmath18 and a choice of an @xmath35-dimensional cone in the subdivision containing @xmath17 ( i.e. , in case there are more than one such cone , we are free to choose any one of them ) .
the factorization problem then is : given two @xmath35-dimensional nonsingular cones @xmath146 and a vector @xmath20 , there exists a nonsingular cone @xmath21 obtained from both @xmath19 and @xmath18 by sequences of star subdivisions along @xmath17 .
it is clear that the factorization problem has a solution only if the interiors of @xmath18 and @xmath19 intersect nontrivially .
we assume that this is the case initially and after every subdivision we choose a cone containing @xmath17 such that this condition again holds .
an extreme case of the factorization problem is when @xmath147 .
then a factorization along any vector @xmath148 ( for instance , @xmath149 with rationally independent coordinates ) is also a factorization along @xmath17 . if @xmath150 , we reduce the factorization problem to the case of @xmath17 with rationally independent coordinates as follows .
the first reduction step is to star subdivide both @xmath19 and @xmath18 along @xmath17 to get to the situation where @xmath151 , @xmath34 and @xmath152 such that the coordinates of @xmath17 with respect to @xmath153 ( hence also with respect to @xmath154 ) are rationally independent . for this write @xmath155 and
consider the vector @xmath156 with nonnegative entries .
it is a simple exercise to show that after a finite sequence of column operations where one subtracts @xmath95 from @xmath157 for @xmath158 , we get to the vector ( after reordering the components ) @xmath159 , such that @xmath160 are linearly independent over @xmath39 . after a similar sequence of star subdivisions of @xmath18
, we get @xmath161 .
note that @xmath162 is the smallest subspace of @xmath163 spanned by rational vectors and containing @xmath17 .
the next step is to use the rationally independent case and factor @xmath164 .
thus after a finite sequence of star subdivisions of @xmath18 and @xmath19 we may assume that the two cones have a common face @xmath165 containing @xmath17 .
after additional subdivisions of @xmath19 we may also assume that @xmath166 .
the final step is to consider the projection @xmath167 , and proceed by induction on dimension the same way as in the proof of theorem [ thm2 ] .
we recall in this section the monomialization theorem of s. d. cutkosky , and then prove theorem [ thm1 ] .
let @xmath168 be a regular local ring of dimension @xmath35 containing a field @xmath9 of characteristic zero , and let @xmath11 be a valuation on the fraction field of @xmath7 , such that the valuation ring @xmath169 dominates @xmath7 .
let @xmath170 be a subset of a system of regular parameters @xmath171 of @xmath7 .
then the homomorphism @xmath172)_p,\ ] ] for some @xmath173 , and @xmath174 a prime ideal lying over @xmath175 , is called a _ monoidal transform _ of @xmath7 .
if @xmath176 is again dominated by the valuation @xmath11 , we say that the monoidal transform is a transform _ along the valuation @xmath11 .
_ geometrically , a monoidal transform is obtained by blowing up a smooth center and localizing at a point @xmath174 above @xmath175 determined by the valuation . in the following , we will be interested in monoidal transforms with @xmath177 .
let @xmath7 and @xmath8 be two excellent regular local rings of dimension @xmath35 containing a field @xmath9 of characteristic zero , both dominated by a valuation @xmath11 on their common fraction field . s. d. cutkosky proved in @xcite that after a sequence of monoidal transforms of @xmath7 and @xmath8 , one can express a system of regular parameters of @xmath8 as monomials in regular parameters of @xmath7 .
more precisely , if @xmath11 has rank @xmath96 and rational rank @xmath35 ( i.e. , the value group can be embedded in @xmath178 and it contains @xmath35 rationally independent elements ) , then after a finite sequence of monoidal transforms , we may assume that a system @xmath179 of regular parameters of @xmath8 can be expressed in terms of regular parameters @xmath180 of @xmath7 as : @xmath181 where @xmath182 are nonnegative integers and @xmath183 .
note that @xmath184 are rationally independent positive real numbers .
if @xmath11 is an arbitrary valuation , then the matrix @xmath185 is block diagonal , with @xmath186 corresponding to the same block having rationally independent values @xmath187 ( @xcite , theorem 4.4 ) . in the following proof
we will perform monoidal transforms with centers @xmath188 or @xmath189 , with @xmath85 and @xmath86 lying in the same block , hence we may assume that @xmath11 has rank @xmath96 and rational rank @xmath35 .
now let us consider the situation of theorem [ thm1 ] .
we assume that an embedding @xmath190 is given by monomials as above , and we have to show that after a sequence of monoidal transforms along @xmath11 , we get ( renaming parameters ) @xmath191 for @xmath192 .
this follows directly from theorem [ thm2 ] , once we express the problem in terms of cones and subdivisions .
let @xmath193 be a nonsingular cone , and let @xmath194 be the cone defined by @xmath195 where @xmath185 is the matrix of exponents above . since @xmath183 , the cone @xmath19 is nonsingular .
we also let @xmath196 be a vector @xmath197 .
now one can easily check that the monoidal transform of @xmath7 with center @xmath198 along @xmath11 corresponds to the star subdivision of @xmath19 at @xmath199 along @xmath17 ( which in terms of the matrix @xmath185 corresponds to adding one column to another ) , and similarly for @xmath8 and @xmath18 ( in terms of @xmath185 , subtract one row from another ) . applying theorem [ thm2 ] , after a finite sequence of monoidal transforms of @xmath7 and @xmath8 ,
the matrix @xmath182 is the identity matrix , hence @xmath200 .
k. matsuki , _ correction : `` a note on the factorization theorem of toric birational maps after morelli and its toroidal extension '' [ tohoku math .
j. ( 2 ) 51 ( 1999 ) , no .
4 , 489537 ] by d. abramovich , matsuki and s. rashid _ , tohoku math . j. ( 2 ) 52 ( 2000 ) , no .
4 , 629631 . | the strong factorization conjecture states that a proper birational map between smooth algebraic varieties over a field of characteristic zero can be factored as a sequence of smooth blowups followed by a sequence of smooth blowdowns .
we prove a local version of the strong factorization conjecture for toric varieties . combining this result with the monomialization theorem of s. d. cutkosky
, we obtain a strong factorization theorem for local rings dominated by a valuation . |
the structure of the dark matter halos of galaxies provides important clues to their formation and dynamical evolution ( e.g. sackett 1996 ; de zeeuw 1996 , 1997 ) .
for example , in the cold dark matter ( cdm ) scenario ( e.g. ostriker 1993 ) there is evidence that the density profiles of halos have a universal form essentially independent of the halo mass or @xmath12 ( navarro , frenk , & white 1997 ; though see moore et al .
the intrinsic shapes of cdm halos are oblate - triaxial with ellipticities similar to the optical isophotes of elliptical galaxies ( e.g. dubinski 1994 ) . the global shape of a halo also has implications for the mass of a central black hole ( e.g. merritt & quinlan 1997 ) .
at present accurate constraints on the intrinsic shapes and density profiles of early - type galaxies are not widely available ( e.g. sackett 1996 ; olling & merrifield 1997 ) . stellar dynamical analyses that have incorporated the information contained in high order moments of stellar velocity profiles
have made important progress in limiting the uncertainty in the radial distribution of gravitating mass arising from velocity dispersion anisotropy ( rix et al .
1997 ; gerhard et al .
1997 ) . however , as indicated by the paucity of such stellar dynamical measurements , the required observations to obtain precise constraints at radii larger than @xmath13 are extensive , and the modeling techniques to recover the phase - space distribution function are complex .
it is also unclear whether this method can provide interesting constraints on the intrinsic shapes since only weak limits on the range of possible shapes have been obtained from analysis of velocity profiles out to @xmath14 @xmath15 ( e.g. statler 1994 ) .
interesting measurements of the ellipticity of the gravitating mass have been obtained for two polar ring galaxies ( sackett et al .
1994 ; sackett & pogge 1995 ) and from statistical averaging of known gravitational lenses ( e.g. keeton , kochanek , & falco 1997 ) , but owing to the rarity of these objects it is possible that the structures of their halos are not representative of most early - type galaxies
. moreover , gravitational lenses , which are biased towards the most massive galaxies , only give relatively crude constraints on the ellipticity and radial mass distribution for any individual system and only on scales similar to the einstein radius ( e.g. kochanek 1991 ) .
the x - ray emission from hot gas in isolated early - type galaxies ( forman , jones , & tucker 1985 ; trinchieri , fabbiano , & canizares 1986 ; for a review see sarazin 1997 ) probably affords the best means for measuring the shapes and radial mass distributions in these systems ( for a review see buote & canizares 1997b ; also see schechter 1987 and the original application to the analogous problem of the shapes of galaxy clusters by binney & strimple 1978 ) .
the isotropic pressure tensor of the hot gas in early - type galaxies greatly simplifies measurement of the mass distribution over stellar dynamical methods .
moreover , since the shape of the volume x - ray emission traces the shape of the gravitational potential independent of the ( typically uncertain ) gas temperature profile ( buote & canizares 1994 , 1996a ) , the shape of the mass distribution can be accurately measured in a way that is quite robust to the possible complicating effects of multi - phase cooling flows and magnetic fields ( see buote & canizares 1997b ) . presently , x - ray measurements of the mass distributions in early - type galaxies are inhibited by limitations in the available data .
the _ rosat _
@xcite position sensitive proportional counter ( pspc ) @xcite has inadequate spatial resolution ( psf @xmath16 fwhm ) to map the detailed mass distributions for all but the largest nearby galaxies , and the limited spectral resolution and band width complicates interpretation of the measured temperature profiles ( buote & canizares 1994 ; trinchieri et al . 1994 ; buote & fabian 1997 ) . although equipped with superior spatial resolution ( psf @xmath17 fwhm ) , the _ rosat _ high resolution imager ( hri ) @xcite has too small an effective area and too large an internal background to provide images of sufficient quality for many galaxies for radii @xmath18 . among the few galaxies with detailed measurements of their radial mass profiles
are ngc 507 ( kim & fabbiano 1995 ) , ngc 1399 ( rangarajan et al . 1995 ; jones et al .
1997 ) , ngc 4472 ( irwin & sarazin 1996 ) , ngc 4636 ( trinchieri et al .
1994 ) , ngc 4649 @xcite , and ngc 5044 ( david et al . 1994 ) .
the shape of the gravitating mass has been measured via x - ray analysis for the e4 galaxy ngc 720 and the e7/s0 galaxy ngc 1332 and found to be at least as elongated as the optical isophotes ( buote & canizares 1994 , 1996a , 1997a ) . for ngc 720 , which has more precise constraints , the ellipticity of the gravitating matter is @xmath19 ( 90% confidence ) compared to the intensity weighted ellipticity of the optical light , @xmath20 ( buote & canizares 1997a ) .
in addition , the x - ray isophotes of ngc 720 twist from being aligned with the optical isophotes within @xmath15 to a position @xmath21 offset at larger radii .
this twist , when combined with the ellipticities of the x - ray isophotes , can not be explained by the projection of a reasonable triaxial matter distribution and thus may implicate a dark matter halo misaligned from the stars ( buote & canizares 1996b ; romanowsky & kochanek 1997 ) .
ngc 720 and ngc 1332 were selected for analysis since they are isolated , significantly elongated in the optical , sufficiently bright , and sufficiently dominated by emission from hot gas in the _ rosat _ band . in this paper
we present x - ray analysis of the classic `` shell '' galaxy , ngc 3923 , which is the last galaxy of which we are aware that satisfies these selection criteria and has deep _ rosat _ observations .
this isolated e4 galaxy has both archival _ rosat _ pspc and hri data and its _ asca _ spectrum has been analyzed previously @xcite .
this will serve as our final case study until the impending launch of _ axaf _ revolutionizes this field .
the organization of this paper is as follows . in [ obs ]
we describe the _ rosat _ observations and the data reduction .
we discuss removal of point sources in [ pt ] .
measurements of the ellipticities of the x - ray isophotes and the radial profiles are described in [ e0 ] and [ radpro ] respectively .
analysis of the pspc spectrum is presented in [ spectra ] .
we give results for the geometric test for dark matter in
[ gt ] and constraints on the shape and radial mass distribution from detailed hydrostatic models in [ models ] . finally , in [ conc ] we give our conclusions .
ngc 3923 was observed with the pspc for 14.5 ks from 17 - 19 december 1991 and for 24 ks from 23 - 26 june 1993 .
both observations were positioned at the field center where the point spread function is smallest . since
spatial resolution is of principal importance for our analysis of the shape of the x - ray surface brightness , we analyze only the pspc data in the hard band ( pi channels 42 - 201 , @xmath22 kev ) to further optimize the size of the psf ; for details regarding the pspc psf see hasinger et al .
( 1993 , updated 1994 ) and see aschenbach for a description of the _ rosat _ x - ray telescope .
we reduced each observation separately with the standard xselect , ftools , and iraf - pros software .
firstly , the events files were cleaned of after - pulse signals by removing any events following within 0.35 ms of a precursor .
we then removed large fluctuations in the light curves indicative of scattered light from the bright earth , sun , or saa ; this resulted in filtered exposures of 13.5 ks for the 1991 observation and 21.8 ks for the 1993 observation . to optimize signal - to - noise ( s / n ) and bin - size requirements for computing ellipticities ( see [ e0 ] ) , we binned the cleaned events files into images with @xmath23 pixels .
exposure maps for the appropriate energy band ( pi=42 - 201 ) were then generated for each image and were then used to create flattened images ; note that the intrinsic resolution of the exposure maps is @xmath24 .
finally , we aligned the images using bright point sources in the field and then combined the images .
the final image is displayed in figure [ fig.n3923 ] ngc 3923 was observed with the hri for 45 ks from 10 - 17 january 1995 and for 26 ks from 22 - 26 june 1995 .
both observations were positioned at the field center where the psf is smallest ( david et al .
for detailed explanation of our reduction of hri data we refer the reader to the related study of ngc 720 by buote & canizares .
we restricted the data to those in pulse - height analyzer ( pha ) bins 2 - 8 since they maximized the s / n of the data of each observation .
for each observation we binned the events into images with @xmath25 pixels to optimize determination of ellipticities within a @xmath26 radius of the galaxy center ( [ e0 ] ) .
exposure maps were generated with the standard software for each image and then used to flatten the images .
an accurate aspect solution is critical to analysis of isophote shapes with the hri on small scales @xmath9 , with the amplitude of the asymmetrical distortion due to incomplete aspect correction typically being most important for @xmath27 @xcite .
the s / n of each of the ngc 3923 observations are too small to usefully perform the aspect correction algorithm of morse .
the low s / n also makes it difficult to make strong statements about the ellipticity of isolated point sources indicative of aspect error . as a result ,
following buote & canizares we search for possible aspect errors by examining the positions of point sources for each of the obis ; i.e. the time intervals during which the spacecraft continuously pointed on the target .
three obis contain most of the `` on time '' of the january observation : 19 ks for obi-6 , 20 ks for obi-11 , and 9 ks for obi-14 ; 28 ks is distributed fairly evenly among 12 other obis . using obi-11 as a reference ,
the bright point sources in the ngc 3923 field are displaced by @xmath28 for obi-6 , and @xmath29 for obi-14 .
these shifts , though statistically significant , are very consistent with the expected aspect uncertainties for a typical observation . to register all of the obis to one coordinate frame
, we binned together those obis that were observed close together in time ; i.e. obis 1 - 5 , 6 - 9 , 10 - 11 , 12 - 13 , and 14 - 15 .
once registered to the same coordinate frame , the images were all added . (
note that the image for each of these obi groups was binned as above . )
the june observation proved to be problematic because none of the 19 obis had a long enough exposure to provide very accurate source positions .
hence , we were unable to perform a reliable test of the aspect solution . in order to make some estimate of the error , we grouped those obis closest together in time : obis 1 - 3 , 4 - 12 , 13 - 16 , and 17 - 19 . for three of these groups , less than @xmath25 shifts
were required , but the obi 4 - 12 group required an @xmath30 shift .
this large shift is questionable since all but one source was too faint to even obtain a centroid measurement using the alignment software in iraf . as it is unclear from considerations of the obis
how accurate is the aspect error for the june observation , we compared the ellipticity and orientations ( computed as described in [ e0 ] ) of the surface brightness with those of the january observation .
unfortunately , we find significant disagreement between the two observations ; note the disagreement occurs whether or not we include the obis 4 - 12 for the june observation . in particular , the position angles determined from the january observation are fully consistent with the pspc data and the optical position angles for @xmath31 , which agrees with ngc 720 and ngc 1332 ( buote & canizares 1996a , b ) ; the ellipticities are also consistent with the pspc data .
the june observations , however , have position angles that differ by over @xmath32 and the ellipticities , particularly for @xmath33 are less than 0.1 as opposed to @xmath34 for the january observation . furthermore , the radial profile ( see [ radpro ] ) of the june observation is significantly flatter for @xmath35 .
thus , the isophote shapes and orientations are inconsistent for the two observations .
the differences point to a serious aspect error in the june observations .
that is , although we would expect aspect error to induce ellipticity in an intrinsically circular source , if the distortion occurs nearly along the minor axis of a moderately elliptical source , it will _ reduce _ the ellipticity of the source , as well as alter the position angle .
the flatter inner radial profile of the june data is also consistent with uncorrected wobble . as a result
, we do not include the june data in our analysis .
( this is not critical since the s / n to be gained over the 45 ks is marginal . )
we display the reduced january observation in figure [ fig.n3923 ] .
it is readily apparent from inspection of figure [ fig.n3923 ] , particularly for the higher s / n pspc data , that the x - ray emission of ngc 3923 is significantly contaminated by foreground / background point sources . within a @xmath36 radius of the galaxy center
at least 8 point sources are easily identified by visual inspection of the pspc image .
( only @xmath37 is shown in the figure . )
we list the positions of these sources in table [ tab.src ] .
.identified point sources [ cols="^,^,^ " , ] 90% confidence values of the gravitating mass corresponding to the oblate and prolate models in table [ tab.shape ] which include 90% uncertainties in the ( isothermal ) temperature .
the statistical errors on the gas mass are less than 10% . in table
[ tab.mass ] we give the spherically averaged masses corresponding to the models in table [ tab.shape ] .
we only quote masses at a few interesting radii to emphasize that without precise constraints on the temperature profile ( and only an aggregate constraint on the isophote shapes ) we do not really constrain the detailed mass profile .
( note that the spectral constraints in [ spectra ] which indicate an approximate isothermal gas apply only for @xmath38 kpc . )
the masses of the @xmath8 and hernquist models agree very well within @xmath39 kpc as expected because over much of this region the large scale lengths of the hernquist model imply an approximate logarithmic slope of -2 . assuming the gas to be isothermal out to @xmath40 kpc , the @xmath8 model , with its flatter slope , has approximately twice the mass of the hernquist model at that distance . the gas mass is less than 1% of the gravitating mass over the entire radius range investigated . using the total @xmath41-band luminosity from faber et al .
scaled to 30 mpc , @xmath42 erg @xmath43 s@xmath44 , we have that the @xmath41-band mass - to - light ratio in solar units is @xmath45 at @xmath39 kpc for the @xmath46 and hernquist models . at @xmath47 kpc ,
the @xmath46 model has @xmath48 while the hernquist model has @xmath49 .
these @xmath50 values within @xmath39 kpc are very consistent with stellar dynamical studies of ellipticals ( e.g. van der marel 1991 ) .
the larger ( extrapolated ) @xmath50 values at @xmath51 kpc agree with previous x - ray studies of ellipticals ( e.g. sarazin 1997 ) indicating that ngc 3923 has a mass distribution typical for a galaxy of its luminosity .
we have analyzed the gravitating matter distribution of the e4 galaxy ngc 3923 using archival x - ray data from the _ rosat _ pspc and hri .
analysis of the pspc data , which allows more precise constraints than the hri data , demonstrates that the x - ray isophotes are significantly elongated with ellipticity @xmath0 ( 90% confidence ) for semi - major axis @xmath1 kpc and have position angles aligned with the optical isophotes within the estimated uncertainties .
a bright point source located @xmath52 along the major axis inhibits reliable ellipticity constraints for larger radii . by applying a `` geometric test '' for dark matter , which essentially compares the shapes of the observed x - ray isophotes to those predicted
if mass traces the optical light @xmath53 ( independent of the poorly constrained temperature profile of the gas ) , we found that the ellipticity of the pspc x - ray surface brightness exceeds that predicted by the constant @xmath54 hypothesis at the 80%-85% confidence level .
the `` geometric test '' result is conservative since it only considers signatures of dark matter that are distributed differently from the optical light .
although the evidence for dark matter from the geometric test is marginal , the results from models which employ an explicit solution of the hydrostatic equation assuming an isothermal gas ( which is supported by the pspc spectrum [ spectra ] ) indicate that @xmath2 is highly inconsistent with the radial profiles of the pspc and hri data ( @xmath55 for 16 dof ) .
this particular discrepancy arises because @xmath53 is too centrally concentrated : the derived scale length of the gravitating matter is approximately 1.5 - 2 times that of @xmath53 .
the ellipticities predicted by this @xmath2 model fall below the pspc data at a significance slightly greater than the 90% level .
modeling the gravitating mass with a density run @xmath8 or with a hernquist profile we find that the ellipticity of the gravitating matter is , @xmath56 ( 90% confidence ) , which is larger than the intensity weighted optical ellipticity @xmath57 .
this evidence for dark matter which is more flattened and more extended than @xmath53 is similar to our conclusions from previous x - ray studies of two other ellipticals , ngc 720 and ngc 1332 , but at somewhat smaller significance level than for ngc 720 ( e.g. buote & canizares 1997b ) .
these results are consistent with analyses of known gravitational lenses ( e.g. keeton , kochanek , & falco 1997 ) , two polar ring galaxies ( sackett et al .
1994 ; sackett & pogge 1995 ) , and flaring disks in spiral galaxies ( e.g , olling 1996 ) .
the ellipticities of the gravitating matter derived from our x - ray analyses and these other methods are consistent with those of halos produced by cdm simulations ( e.g. dubinski 1994 ) .
if an isothermal gas is assumed then models with matter density @xmath8 are favored over hernquist models ( and similar models like the universal cdm profile of navarro et al .
for @xmath58 the @xmath8 model marginally fits the data better than the hernquist model .
however , most of the difference in these models occurs in the central radial bins where the effects of multi - phase cooling flows , magnetic fields , and discrete sources could affect the surface brightness profiles , though we have argued the effects are unlikely to be important ( see [ models ] ) .
( the derived shape of the gravitating mass is mostly robust to these issues buote & canizares 1997b . )
this support for nearly @xmath59 profiles agrees with previous studies of gravitational lenses ( e.g. maoz & rix 1993 ; kochanek 1995 ) , although a recent paper finds that density profiles with changing slopes ( e.g. hernquist and nfw ) are preferred @xcite .
an emission component that is proportional to @xmath53 can not contribute significantly to the _ rosat _ x - ray emission of ngc 3923 , and thus discrete sources should not affect our constraints on the gravitating matter ( buote & canizares 1997a ) .
however , the _ asca _ spectral data when fitted with two thermal components yield a cold component , @xmath60 kev , and a hot component , @xmath61 kev , where the relative flux of cold - to - hot is @xmath62 in the _ rosat _ band @xcite . the conventional interpretation of the hot component ( e.g. matsumoto et al . 1997 ; loewenstein & mushotzky 1997 ) is that it arises from discrete sources .
but our analysis ( [ radpro ] ) shows that @xmath63 of the 0.5 - 2 kev emission can not be distributed like the optical light which would be expected of discrete sources .
hence , either the emission from discrete sources is not distributed like @xmath53 , or the hot component obtained from the spectral fits can not be entirely due to discrete sources as suggested by buote & fabian .
the constraints we have obtained for ngc 720 , ngc 1332 , and now ngc 3923 from analyses of their x - ray isophote shapes and radial surface brightness profiles provide an initial demonstration of the power of x - ray analysis for probing the shape and radial distribution of gravitating matter in early - type galaxies .
the next generation of x - ray satellites , particularly _ axaf _ and _ xmm _ , have the capability to accurately map x - ray isophote shapes and orientations from the cores ( @xmath64 ) out to 10s of kpc for many galaxies ) which hindered the present analysis of ngc 3923 . ] . the spatially resolved spectra provided by these future missions will allow more precise constraints on temperature gradients and the contribution from discrete sources .
thus , unlike most other methods , obtaining interesting x - ray constraints on the shape and radial density profile of the gravitating matter will be possible for a large sample of early - type galaxies since the x - ray analysis is applicable to any isolated early - type galaxy whose soft x - ray emission ( @xmath65 kev ) is dominated by hot gas .
allen s. w. , fabian a. c. , kneib j .-
, 1996 , mnras , 279 , 615 anders e. , grevesse n. , 1989 , geochimica et cosmochimica acta , 53 , 197 arnaud k. , 1996 , in jacoby g. and barnes j. , eds . , astronomical data analysis software and systems v , asp conf .
series volume 101 , p17 aschenbach b. , 1988 , appl .
opt . , 27(8 ) , 1404 baluciska - church m. , mccammon d. , 1992 , 400 , 699 binney j. , strimple o. , 1978 , mnras , 185 , 473 binney j. , tremaine s. , 1987 , galactic dynamics .
princeton univ . press , princeton birkinshaw m. , davies r. l. , 1985 , apj , 291 , 32 brighenti f. , mathews w. g. , 1997 , apj , 486 , l83 broeils a. , 1997,in d. zaritsky ed . , uc santa cruz workshop on galactic halos , asp conf .
series , in press buote d. a. , canizares c. r. , 1994 , apj , 427 , 86 buote d. a. , canizares c. r. , 1996a , apj , 457 , 177 buote d. a. , canizares c. r. , 1996b , apj , 468 , 184 buote d. a. , canizares c. r. , 1997a , apj , 474 , 650 buote d. a. , canizares c. r. , 1997b , in d. zaritsky ed . ,
uc santa cruz workshop on galactic halos , asp conf .
series , in press ( astro - ph/9710001 ) buote d. a. , fabian a. c. , 1998 , mnras , in press , ( astro - ph/9707117 ) buote d. a. , tsai j. c. , 1995 , apj , 439 , 29 buote d. a. , tsai j. c. , 1996 , apj , 458 , 27 canizares c. r. , fabbiano g. , trinchieri g. , 1987 , apj , 312 , 503 carter d. , metcalfe n. , 1980 , mnras , 191 , 325 cavaliere a. , fusco - femiano r. , 1976 , a&a , 49 , 137 david l. p. , jones c. , forman w. , daines s. , 1994 , apj , 428 , 544 david l. p. , et al .
, 1997 , the rosat high resolution imager ( hri ) calibration report , ( u.s .
rosat science data center
smithsonian astrophysical observatory ) de zeeuw p. t. , 1996 , in lahav o. , terlevich e. , terlevich r. j. eds .
, gravitational dynamics , cambridge contemporary astrophysics , ( cambridge university press : cambridge ) de zeeuw p. t. , 1997 , in arnaboldi m. , da costa g. s. , p. saha , eds . , the nature of elliptical galaxies , proceedings of the second stromlo symposium , asp conference series , in press ( astro - ph/9704095 ) dubinski j. , 1994 , apj , 431 , 617 faber s. m. , wegner g. , burstein d. , davies r. l. , dressler a. , lynden - bell d. , terlevich r. j. , 1989 , apjs , 69 , 763 fabricant d. , rybicki g. , gorenstein p. , 1984
, apj , 286 , 186 forman w. , jones c. , tucker w. , apj , 1985 , 293 , 102 gerhard o. e. , jeske g. , saglia r. p. , bender r. , 1997 , mnras , in press ( astro - ph/9710129 ) hasinger g. , turner t. j. , george i. m. , boese g. , 1993 , mpe / ogip calibration memo cal / ros/93 - 015 hasinger g. , burg r. , giacconi r. , hartner g. , schmidt m. , trmper j. , zamorani g. 1993 , a&a , 275 , 1 hernquist , l. 1990 , apj , 356 , 359 irwin j. a. , sarazin c. l. , 1996 , apj , 471 , 683 jedrzejewski r. , 1987 , mnras , 226 , 747 jones c. , stern c. , forman w. , breen j. , david l. , tucker w. , franx m. , 1997 , apj , 482 , 143 kaastra j. s. , mewe r. , 1993 , a&as , 97 , 443 keeton c. , kochanek c. s. , falco e. , 1997 , apj , submitted ( astro - ph/9708161 ) kim d .- w .
, fabbiano g. , 1995 , apj , 441 , 182 kim d .- w .
, fabbiano g. , trinchieri g. , 1992 , apj , 393 , 134 kley w. , mathews , w. g. , 1995 , apj , 438 , 100 kochanek c. s. , 1991 , apj , 373 , 354 kochanek c. s. , 1995 , apj , 445 , 559 liedahl d. a. , osterheld a. l. , goldstein w. h. , 1995 , apj , 438 , l115 loewenstein m. , mushotzky r. f. , 1997 , in d. zaritsky ed . , uc santa cruz workshop on galactic halos , asp conf .
series , in press ( astro - ph/9710339 ) maoz d. , rix h .- w .
, 1993 , apj , 416 , 425 matsumoto h. , koyama k. , awaki h. , tsuru t. , loewenstein m. , matsushita k. , 1997 , apj , 482 , 133 merritt d. , quinlan g. d. , 1997 , apj , submitted ( astro - ph/9709106 ) mewe r. , gronenschild e. h. b. m. , van den oord g. h. j. , 1985 , a&as , 62 , 197 milgrom m. , 1983 , apj , 270 , 365 moore b. , governato f. , quinn t. , stadel j. , lake g. 1997 , apj , submitted ( astro - ph/9709051 ) morse j. a. , 1995 , in e. m. schlegel r. petre eds . , the soft x - ray cosmos ( aip conf . series 313 ) , ( aip press : new york ) , 252 navarro j. f. , frenk c. s. , white s. d. m. , 1997 , mnras , in press nulsen p. e. j. , stewart g. c. , fabian a. c. , 1984 , mnras , 208 , 185 olling r. p. , 1996
, aj , 112 , 457 olling r. p. , merrifield m. r. , in d. zaritsky ed . ,
uc santa cruz workshop on galactic halos , asp conf .
series , in press ( astro - ph/9710224 ) ostriker j. p. , 1993 ,
ara&a , 31 , 689 pfeffermann e. , et al . , 1987 , proc .
spie , 733 , 519 rangarajan f. v. n. , fabian a. c. , forman w. r. , jones c. , 1995 , mnras , 272 , 665 raymond j. c. , smith b. w. , 1977 , apjs , 35 , 419 rix h .- w .
, de zeeuw p. t. , cretton n. , van der marel r. p. , carollo c. m. , 1997 , apj , 488 , 702 romanowsky a. j. , kochanek c. s. , 1997 , apj , in press ( astro - ph/9708212 ) sackett p. d. , rix h .- w . , jarvis b. , freeman k. c. , 1994 , apj , 436 , 629 sackett p. d. , pogge r. w. , 1995 , in e. m. schlegel r. petre eds . , the soft x - ray cosmos ( aip conf .
series 313 ) , ( aip press : new york ) , 141 sackett p. d. , 1996 , in c. s. kochanek , j. n. hewitt eds . ,
gravitational lensing ( iau 173 ) , ( dordrecht : reidel ) , ( astro - ph/9508098 ) sarazin c. l. , 1997 , in arnaboldi m. , da costa g. s. , and p. saha , eds . ,
the nature of elliptical galaxies , proceedings of the second stromlo symposium , in press schechter , p. l. 1987 , in t. de zeeuw ed . , structure and
dynamics of elliptical galaxies ( iau symp .
127 ) , ( dordrecht : reidel ) , 217 stark a. a. , gammie c. f. , wilson r. w. , bally j. , linke r. a. , heiles c. , hurwitz m. 1992 , apjs , 79 , 77 statler t. s. , 1994 , aj , 108 , 111 strimple o. , binney j. , 1979 , mnras , 185 , 473 tanaka , y. , inoue h. , holt s. s. 1994 , pasj , 46 , l37 thomas p. a. , fabian a. c. , nulsen p. e. j. , 1987 , mnras , 228 , 973 trinchieri g. , fabbiano g. , canizares c. , 1986 , apj , 310 , 637 trinchieri g. , kim d .- w .
, fabbiano g. , canizares c. , 1994 , apj , 428 , 555 trmper j. , 1983 , adv . space res . , 2 , 241 van der marel , r. p. , 1991 ,
mnras , 253 , 710 williams l. r. , 1997 , mnras , in press ( astro - ph/9709283 ) | we present analysis of the shape and radial mass distribution of the e4 galaxy ngc 3923 using archival x - ray data from the _ rosat _ pspc and hri .
the x - ray isophotes are significantly elongated with ellipticity @xmath0 ( 90% confidence ) for semi - major axis @xmath1 kpc and have position angles aligned with the optical isophotes within the estimated uncertainties . applying the geometric test for dark matter , which is independent of the gas temperature profile ,
we find that the ellipticities of the pspc isophotes exceed those predicted if @xmath2 at a marginal significance level of @xmath3 for oblate ( prolate ) symmetry . detailed hydrostatic models of an isothermal gas yield ellipticities for the gravitating matter , @xmath4 ( 90% confidence ) , which exceed the intensity weighted ellipticity of the @xmath5-band optical light , @xmath6 ( @xmath7 ) .
we conclude that mass density profiles with @xmath8 are favored over steeper profiles if the gas is essentially isothermal ( which is suggested by the pspc spectrum ) and the surface brightness in the central regions @xmath9 is not modified substantially by a multi - phase cooling flow , magnetic fields , or discrete sources .
we argue that these effects are unlikely to be important for ngc 3923 .
( the derived @xmath10 range is very insensitive to these issues . )
our spatial analysis also indicates that the allowed contribution to the _ rosat _ emission from a population of discrete sources with @xmath11 is significantly less than that indicated by the hard spectral component measured by _
asca_.
galaxies : elliptical and lenticular , cd galaxies : fundamental parameters galaxies : individual ( ngc 3923 ) galaxies : structure x - rays : galaxies . |
strong correlations among the electrons and orbital degeneracy both play a major role in the low temperature physics of transition metal compounds @xcite . materials within this class range from the recently discovered superconducting iron pnictides to the manganites showing the colossal magnetoresistance . due to the short spatial extension of the 3d - orbitals in the transition metal atoms and the typical crystal structure of these compounds , screening of the coulomb interaction can be considered weak , leading to the above mentioned strong local electron - electron interactions . within the group of transition metal compounds ,
the cubic perovskite structure is a particularly common one . in this structure
the 5-fold degenerate d - orbitals split into three - fold degenerate @xmath0-orbitals and two - fold degenerate @xmath1-orbitals .
therefore , besides strong correlations also orbital degeneracy will play an important role in these compounds .
especially near integer filling , orbital degeneracy can induce long - range orbital order @xcite , for which the expectation value to find an electron in one orbital depends on the lattice site and the orbital .
it can be accompanied by a lattice distortion , jahn - teller distortion , caused by the coupling between electrons and the lattice @xcite . however , even without such a lattice distortion it can be energetically favorable to form an orbitally ordered state @xcite .
although the perovskite crystal structure is comparatively simple , a full investigation of material - specific properties is still a big challenge . on the other hand , as a rather large number of different compounds
do show qualitatively similar physics , it is suggestive to understand these common aspects by studying a model concentrating on the most important ingredients of the 3d transition metal perovskites . as has been discussed by a variety of authors ,
such a model is the multi - orbital hubbard hamiltonian @xcite , which will therefore build the basis of our investigations . in this article
we especially want to focus on the interplay between orbital degeneracy and strong correlations leading to either a competition or cooperation between long - range ordered phases of spin , orbital and charge degrees of freedom . particularly interesting for manganites
is the case of a twofold degenerate d - band at the fermi energy , and here the special point of quarter filling seems to play a major role for the physics of this class of compounds .
therefore , our aim is to study the physics of the two - orbital hubbard hamiltonian at quarter filling , with special emphasis laid on the phase diagram for the magnetic and the orbital order , and the changes in various physical quantities across the phase boundaries .
this article is organized as follows .
after this introduction we will specify the model , shortly explain the used methods , and give a short overview about the ground state properties at quarter filling .
thereafter we will study the influence of temperature and of magnetic fields on the ordered phases . as a particularly interesting quantity with respect to experiment we will also present results for the conductivity and its changes across the phase boundaries .
a summary will conclude the paper .
as already noted in the introduction , a reasonable qualitative description of the low - energy properties of 3d - transition metal perovskites can be obtained by the multi - orbital hubbard model @xcite @xmath2 where @xmath3 creates an electron at site @xmath4 , with spin @xmath5 in orbital @xmath6 .
furthermore , @xmath7 is the density operator and @xmath8 is the spin operator for the electrons .
@xmath9 corresponds to a hopping of the electrons between nearest neighbor sites and @xmath10 represents a pure local two - particle interaction .
the interaction consists of an intra - orbital density - density interaction @xmath11 , an inter - orbital density - density interaction @xmath12 , as well as a ferromagnetic hund s coupling between the orbitals , @xmath13 .
we here neglect the pair - hopping term in the hamiltonian , which should have only minor quantitative and no qualitative influence as we perform calculations for strong repulsive @xmath11 and away from half filling .
nevertheless , it should be stated that there is no rotational orbital symmetry due to the exclusion of the pair - hopping term . however , due to this approximation , it is possible to include the orbital occupation as conserved quantum number into the calculations , which considerably simplifies and speeds up the numerical calculations . to check the validity of this approximation
, we have performed a few additional calculations including the pair - hopping term , but no significant differences have been found .
although including only local interactions and nearest neighbor hopping terms , the hubbard model is very challenging . for calculating the magnetic and orbital phase diagrams , we use the dynamical mean field theory ( dmft ) @xcite . capturing the local physics correctly ,
it has proved to be a very powerful instrument for analyzing and understanding strong correlation effects . moreover , although for long - range order it is closely connected to standard mean field theory , the inclusion of local dynamical properties significantly renormalizes the physical properties , and even completely suppresses ordering where static mean field approaches would predict some . therefore , even if we can not account for spatial fluctuations properly , the dmft results will give a reasonable qualitative and thermodynamically consistent account of possible phases and becomes exact in the limit of infinite spatial dimensions . as we are in this study interested in the fundamental aspects of the interplay between orbital degeneracy and strong correlations , dmft is well suited . for a realistic comparison to transition metal oxides ,
e.g. manganites , the lattice structure and also phonon modes should be taken into account .
the remaining obstacle is to solve the dmft self - consistency equations , which are tantamount to calculate spectral functions for a multi - orbital single impurity anderson model @xcite .
as we want to concentrate on @xmath1-type systems with a two - fold degenerate ground state multiplet , we can employ the numerical renormalization group ( nrg ) @xcite , which is able to reliably calculate spectral functions at low and zero temperature also including spin- and orbital - symmetry - broken states . to be able to calculate spectral functions at arbitrary temperatures within nrg
, we use the complete fock space algorithm @xcite .
an important aspect of the dmft is that the lattice structure enters only through the non - interacting density of states ( dos ) .
furthermore , apart from quantitative details , the basic physical properties are rather insensitive to the actual form of the dos , as long as particle - hole symmetry holds . as we are interested in qualitative aspects of the physics of the two - orbital hubbard model , we therefore have a certain freedom to choose a numerically convenient dos .
we thus choose a semi - elliptic local density of states with bandwidth @xmath14 for the non - interacting system .
for the nrg calculations we use @xmath15 to @xmath16 states kept per nrg iteration and an nrg discretization parameter @xmath17 .
@xcite throughout this article the local hubbard interaction is set to @xmath18 , which is a reasonable value for transition metal oxides .
ferromagnetism and orbital order in a multi - orbital hubbard model in infinite dimensions was analyzed by other authors before @xcite .
however , up to now the connection between the itinerant ferromagnetic phase , most pronounced for a filling @xmath19 , and the orbitally ordered ferromagnetic phase at quarter filling was not sufficiently analyzed .
the ground state phase diagram of the two - orbital hubbard model was recently analyzed in some detail by @xcite .
in particular , for quarter filling the schematic phase diagram shown in fig . [ groundstate_quarter ] was obtained . in this figure both interaction parameters @xmath20 and the hund s coupling @xmath21 are treated as independent parameters .
the black line represents the relation @xmath22 corresponding to the orbital symmetric case , in which the atomic ground state at half filling is a spin triplet . .
the black line represents @xmath22 .
[ groundstate_quarter ] ] the ground state phase diagram consists of 4 different phases . for the case of both interaction parameters ( @xmath20 and @xmath21 ) being small , there exists only a paramagnetic metallic phase . for small hund s coupling @xmath21 but large repulsive inter - orbital density - density interaction @xmath20
, there appears an antiferro - orbitally ordered phase without spin polarization : like for the nel state of the antiferromagnet , the lattice exhibits a bipartite ab - structure , where electrons on the a - sites of the lattice only occupy the orbital @xmath23 , while on the b - sites of the lattice they only occupy the orbital @xmath24 . if , on the other hand , @xmath25 is kept small and the hund s coupling @xmath21 is increased , one encounters for @xmath26 a homogeneous ferromagnetic phase without orbital polarization .
this ferromagnetic phase extends to a filling of approximately @xmath27.@xcite the bulk part of the phase diagram at quarter filling with @xmath28 and @xmath29 is , however , an anti - ferro orbitally ordered , ferromagnetic state .
thus , both types of order are quite obviously supporting each other .
we should note here that for attractive inter - orbital interaction @xmath30 the energetically lowest state is a charge - density wave @xcite , which was however excluded in the phase - diagram fig .
[ groundstate_quarter ] . while the orbitally ordered state at @xmath31 always results in an insulating state , the homogeneous ferromagnetic state and the paramagnetic state both are metallic .
we thus encounter a metal insulator transition ( mit ) at @xmath31 when changing the interaction parameters .
it is well known that strong correlations tend to localize electrons at integer fillings @xcite and that the transition point in the two - orbital hubbard model depends on the strength of the hund s coupling @xmath21 .
@xcite furthermore , the localized electrons in the insulating region tend to form long range order . at quarter filling for the two - orbital hubbard model on a bipartite lattice , this can be most efficiently be done by forming an orbital ordered ferromagnetic state , thus decreasing the energy of the system .
comparing to paramagnetic results , the paramagnetic mit occurs approximately at the same value of hund s coupling as this ferromagnetic mit .
however , the temperature scale of the paramagnetic mit will be smaller , i.e. the paramagnetic mit will be covered by ferromagnetic states having the lower energy for a bipartite lattice . if this mit will be observable in experiments
is at least doubtful for two reasons .
firstly , it is not clear how one can experimentally change the inter - orbital interactions @xmath21 and @xmath20 without also significantly changing the hubbard interaction @xmath11 .
secondly , the neglected coupling to the lattice , which enhances the tendency towards orbital order due to jahn - teller - distortions , will quite likely cover this purely electronic effect . nevertheless , it is important to realize that orbital order can be stabilized without resorting to lattice degrees of freedom by correlation effects only .
: upper panel ) and orbital polarization ( @xmath32 : lower panel ) for different temperatures and hund s coupling at quarter filling and @xmath22 , @xmath33 .
the dip at @xmath34 in the magnetic polarization corresponds to the transition between the orbitally ordered ferromagnet for @xmath35 and the homogeneous ferromagnet for @xmath36 .
[ magnetic_polar],title="fig : " ] : upper panel ) and orbital polarization ( @xmath32 : lower panel ) for different temperatures and hund s coupling at quarter filling and @xmath22 , @xmath33 .
the dip at @xmath34 in the magnetic polarization corresponds to the transition between the orbitally ordered ferromagnet for @xmath35 and the homogeneous ferromagnet for @xmath36 .
[ magnetic_polar],title="fig : " ] experimentally more relevant than changing the interaction parameters is the dependence of physical properties on temperature and magnetic field . in order to limit the number of parameters in our system
, we will from now on focus on the special combination @xmath22 for the interaction parameters , representing the so(3 ) symmetry for isolated atoms @xcite .
figure [ magnetic_polar ] shows the results for the spin and orbital polarization as function of temperature @xmath37 and hund s coupling @xmath21 in a false color plot .
as we have only a limited number of actual data available , we used a polynomial fit to obtain a reasonably smooth density plot .
thus , the precise values and details of the transition lines should be regarded with some caution .
however , the temperature gap in the spin polarization at @xmath34 is real and corresponds to the transition between orbitally ordered ferromagnetism for @xmath35 and homogeneous ferromagnetism ( without orbital order ) for @xmath36 . in the former case ,
the ferromagnetic state is stabilized by orbital order induced by strong @xmath20 , which is an insulator and consequently leads to the formation of local moments , which then order due to double exchange mechanism @xcite . for @xmath36
the orbital order has vanished ( see lower panel of fig .
[ magnetic_polar ] ) , but one still finds a ferromagnetic solution , however with reduced curie temperature and magnetic moment .
the magnetic state in this parameter region remains metallic and must therefore be characterized as itinerant ferromagnet stabilized mainly due to a gain in kinetic energy .
these two rather different ferromagnetic phases are separated by a small region with a paramagnetic and orbitally disordered metal . from fig .
[ magnetic_polar ] we have deduced the schematic phase diagram presented in fig . [ tj_phase ] .
the black points in the graph correspond to the parameters , at which actual calculations have been performed . ,
the black points denote the parameters at which calculations have been performed and from which the phase boundaries were drawn .
[ tj_phase ] ] the orbital order is mainly driven by the inter - orbital density - density interaction @xmath12 . for small hund s coupling ,
meaning strong @xmath20 , the transition temperature should behave like @xmath38 in leading order from perturbation theory due to the same arguments as for the antiferromagnetic nel state at half filling for large @xmath11 .
increasing hund s coupling leads to a decreasing @xmath20 and vanishing orbital order . for @xmath28 and low temperatures
the orbital order comes along with ferromagnetic order .
, @xmath39 . for decreasing temperatures a gap at the fermi - energy @xmath40
is formed , and the spectral functions for @xmath23 and @xmath24 differ from each other . for temperatures below @xmath41
the orbitally ordered ferromagnetic state is stabilized .
the inset in lowest panel shows a magnification around the fermi energy clarifying the orbital order for this temperature .
right panels : spectral functions for @xmath42 .
there is no difference in the spectral functions between @xmath23 and @xmath24 .
additionally , the ferromagnetic transition temperature is much lower .
[ spectral ] ] up to now we have only discussed the static properties in the different phases , frequently referring to one case as an insulator , the other as a metal . to explain this distinction we discuss now the spectral functions for generic parameter values in the different cases .
figure [ spectral ] shows spectral functions corresponding to the cases @xmath35 ( left panels ) and @xmath36 ( right panels ) .
as already observed for the order parameters , one first finds orbital order with decreasing temperature for @xmath35 , and a second transition to a ferromagnet for even lower temperatures . in the case @xmath36 , on the other hand , there is no sign for orbital ordering , but a clear ferromagnetic polarization of the spectra for the lowest shown temperature .
the spectral functions of both systems differ very strongly .
the orbitally ordered system exhibits characteristic van hove singularities due to the reduced translational symmetry , and a gap at the fermi energy @xmath40 , thus being an insulator . the system without orbital order does not show this gap , but a smooth and finite dos at the fermi energy , thus being metallic . however , the insulating behavior for @xmath35 is not only due to orbital order , resulting in a doubling of the unit cell .
the gap can also be found above the critical temperature of the phase transition , see fig .
[ specpmit ] .
this figure shows spectral functions for @xmath43 in the paramagnetic phase for different values of the hund s coupling . by increasing the hund s coupling the gap is closed and a broad peak is formed . at this temperature
, there is no sharp phase transition between the insulator and the metal , but a crossover between both phases .
one can argue that the gap is formed by the strong inter - orbital density - density interaction .
this emphasizes the importance of taking into account the orbital level structure of transition metal oxides when modeling their properties , like the well - known metal insulator transition in v@xmath44o@xmath45 @xcite . , @xmath43 , @xmath22 , and different @xmath21 .
for all shown spectral functions the system is in a paramagnetic state . for increasing @xmath21 , the gap at the fermi - energy @xmath40 vanishes .
[ specpmit ] ]
, @xmath46 , @xmath22 .
the upper two panels show the magnetic and orbital polarization for different temperatures @xmath47 and magnetic fields @xmath48 .
the lower panels show the conductivity @xmath5 scaled by @xmath49 ( compare @xmath50 with fig .
[ magnetic_polarb038 ] ) for @xmath51 different temperatures ( lines are meant as guide to the eye ) , and the spectral functions at @xmath52 .
[ magnetic_polarb034],title="fig : " ] , @xmath46 , @xmath22 .
the upper two panels show the magnetic and orbital polarization for different temperatures @xmath47 and magnetic fields @xmath48 .
the lower panels show the conductivity @xmath5 scaled by @xmath49 ( compare @xmath50 with fig .
[ magnetic_polarb038 ] ) for @xmath51 different temperatures ( lines are meant as guide to the eye ) , and the spectral functions at @xmath52 .
[ magnetic_polarb034],title="fig : " ] , @xmath53 , @xmath22 .
the upper two panels show the magnetic and orbital polarization for different temperatures @xmath47 and magnetic fields @xmath48 .
the lower panels show the conductivity @xmath5 scaled by @xmath54 for @xmath51 different temperatures ( lines are meant as guide to the eye ) , and the spectral functions at @xmath55 .
[ magnetic_polarb038],title="fig : " ] , @xmath53 , @xmath22 .
the upper two panels show the magnetic and orbital polarization for different temperatures @xmath47 and magnetic fields @xmath48 .
the lower panels show the conductivity @xmath5 scaled by @xmath54 for @xmath51 different temperatures ( lines are meant as guide to the eye ) , and the spectral functions at @xmath55 .
[ magnetic_polarb038],title="fig : " ] besides temperature , also the magnetic field can be varied very easily in experiments .
an increasing magnetic field will enhance the magnetic polarization of the system .
however , especially in the vicinity of the phase transition @xmath34 , where the orbitally ordered ferromagnetic state has disappeared , it is a priori not clear how the ferromagnetic state will look like . when applying a magnetic field within the paramagnetic phase between both ferromagnetic phases , the system can form either an orbitally ordered or an orbitally homogeneous state . both states do considerably differ in their physical properties , as the orbitally ordered state is gapped at the fermi energy , while the homogeneous state is not . in figs .
[ magnetic_polarb034 ] and [ magnetic_polarb038 ] we analyze the conductivity of the system for different temperatures and magnetic fields .
the conductivity can be easily calculated within the dmft , as the self - energy is purely local .
we here must take care of a possible ab - ordering ( nel order ) in the orbital index .
the conductivity at @xmath40 can be written as @xcite @xmath56\nonumber\\ a^{\mathcal{s}}_{m,\sigma}(\epsilon,\omega^\prime)&=&-\frac{1}{\pi}\mathfrak{im}\bigl(\frac{\zeta^{\mathcal{\bar s}}_{m,\sigma}}{\zeta^{\mathcal{a}}_{m,\sigma } \zeta^{\mathcal{b}}_{m,\sigma}-\epsilon^2}\bigr)\nonumber\\ b_{m,\sigma}(\epsilon,\omega^\prime)&=&-\frac{1}{\pi}\mathfrak{im}\bigl(\frac{\epsilon}{\zeta^{\mathcal{a}}_{m,\sigma } \zeta^{\mathcal{b}}_{m,\sigma}-\epsilon^2}\bigr)\nonumber\\ \zeta^{\mathcal{s}}_{m,\sigma}&=&\omega^\prime+\mu-\sigma_{m,\sigma}^{\mathcal{s}}(\omega^\prime)\nonumber,\end{aligned}\ ] ] where @xmath57 and @xmath58 are the diagonal and off - diagonal green s functions of the ab - lattice respectively , @xmath59 is the fermi - function , @xmath60 the chemical potential , @xmath61 the local self - energy .
@xmath62 corresponds to the sub - lattice , for which @xmath63 holds .
finally , @xmath64 is the averaged squared fermi - velocity , in which the lattice structure enters @xcite .
the prefactor @xmath65 consists of the resistance quantum and a non - universal part , depending on the details of the lattice . in the following , we use c=1 for convenience .
figure [ magnetic_polarb034 ] summarizes the properties of the system for @xmath46 .
the upper two panels of the figure show the magnetic and orbital polarization , while the lower panel shows conductivity and spectral functions for different temperatures @xmath37 and magnetic fields @xmath66 .
these parameters correspond to the minimum of the curie - temperature ( dip ) in fig .
[ tj_phase ] , where there is neither orbital nor magnetic order for @xmath67 .
we discuss the conductivity instead of the resistivity here as we expect a mit at some point , leading to a drop to zero in @xmath5 , which is easier to visualize than a divergence in @xmath68 .
if one applies a magnetic field at low temperatures , @xmath69 , both spin and orbital polarizations are induced .
assuming a bandwidth of @xmath70 , this corresponds to a temperature of @xmath71 .
the system can still gain more energy by localizing the electrons on different orbitals in the presence of a magnetic field .
this magnetoresistance effect corresponds to a mit triggered by applying a magnetic field .
furthermore , these results agree with the fact that applying a magnetic field at low temperatures near a mit in a one - band model favors the insulating solution @xcite .
looking at the spectral functions for @xmath72 , the middle of the lower panels in the figure , a spin polarized state has formed .
nevertheless , all spectral functions still have spectral weight at the fermi energy .
the majority - spin spectral functions form a two peak structure , firstly even increasing the weight at the fermi energy . further increasing the magnetic field above @xmath73 ( right part of lower panel )
stabilizes the usual orbitally ordered ferromagnetic state with a gap at the fermi energy .
this behavior is reflected in the conductivity , which directly depends on the spectral weight at the fermi energy .
with increasing magnetic field the conductivity initially increases at low temperatures , but then drops to zero as the gap opens .
if one further decreases the temperature , the initial conductivity increase for small magnetic fields will be more pronounced , but it again eventually drops to zero for magnetic fields of approximately @xmath74 .
thus , at low temperatures , one observes a rather dramatic magnetoresistance effect as function of magnetic field , reminiscent of the colossal magnetoresistance effect in the manganites . for higher temperatures the situation is different .
in this case , no orbital polarization is stabilized , and no gap is formed at the fermi energy .
increasing the magnetic field causes the conductivity to increase smoothly , as the spectral weight of the majority spin at the fermi energy is increased .
let us compare these results with those for @xmath53 collected in fig .
[ magnetic_polarb038 ] . for this value of @xmath21
, the orbital order has its highest transition temperature .
nevertheless , for temperatures of approximately the transition temperature , the spectral function is non - zero at the fermi energy , even in the orbitally ordered state . to compare the conductivity in figs .
[ magnetic_polarb034 ] and [ magnetic_polarb038 ] properly , one should also compare the scaling constants @xmath50 .
increasing @xmath21 leads to a more pronounced peak at the fermi energy in the spectral function .
this is reflected in @xmath50 , which for @xmath46 takes the value @xmath49 ( in arbitrary units , neglecting the constant factors @xmath75 for calculating the conductivity , which is the same in both cases ) , while for the orbitally ordered ferromagnetic ground state the scaling constant is @xmath54 .
increasing the temperature without applying a magnetic field , leads first to a vanishing spin polarization and later , at temperatures @xmath76 , to a paramagnetic state . for @xmath77 the system is gapped and the conductivity is nearly zero .
approaching now the transition temperature of the orbitally ordered phase , spectral weight is shifted to the fermi energy . at this point , a magnetic field leads again to an increase of the gap width and to a stabilization of the orbital order .
the corresponding temperature for the magnetoresistance - effect is @xmath78 , again assuming a bandwidth @xmath70 .
this behavior can be seen in the upper panels of fig .
[ magnetic_polarb038 ] in which the transition temperature of the orbital order increases with increasing magnetic field , and in the lower panels directly for the spectral functions or for the conductivity . for @xmath76 ,
the latter now again shows a maximum at some small magnetic field , decreasing again with increasing magnetic field .
however , compared to the case @xmath79 , the effect here is less dramatic . for even higher temperatures , neither an orbital polarization nor a gap at the fermi energy
is induced by a magnetic field resulting in a nearly constant conductivity .
one can do similar calculations for smaller hund s coupling @xmath80 or larger @xmath81 . for smaller hund s coupling
the system is in an insulating state even for temperatures above the critical temperature for the ordered phases .
applying magnetic fields will also increase the transition temperature of the orbitally ordered phase , but as the system is already gapped in the paramagnetic state , the conductivity will not change significantly . for
larger hund s coupling , applying a magnetic field will not induce orbital order anymore .
the system will stay in a metallic state with a correspondingly smooth and weak variation of the physical properties .
we have analyzed the properties of the two - orbital hubbard model at quarter filling for a range of temperatures and magnetic fields .
as expected , ferromagnetic order is prevalent at quarter filling , but depending on the ratio of inter - orbital interaction and hund s exchange , it can appear either without or with accompanying orbital order .
our calculations show that orbital order can be stabilized without including jahn - teller distortions,@xcite and it is stabilized by a strong inter - orbital density - density interaction .
if one decreases the strength of this interaction , the orbital order will eventually vanish . as a result
a metal insulator transition can be found , which can also be triggered by applying a magnetic field close to the transition .
this gives rise to a very strong magnetoresistance effect , as one can drive the system from a metallic into an insulating state .
the strength of this magnetoresistance effect depends sensitively on the hund s coupling @xmath21 .
in particular , we found a critical regime for @xmath21 between an orbitally ordered ferromagnet and an orbitally homogeneous one , where a particularly strong magnetoresistance effect exists .
note that these interaction values for this particularly interesting behavior , @xmath18 and @xmath46 , are not completely unrealistic for transition metal oxides .
it therefore might be interesting to look for @xmath1 systems at quarter filling which are in this parameter regime if they exhibit strong magnetoresistance .
another rather amusing fact is that we found an insulating paramagnetic state for strong inter - orbital density - density interaction reaching to temperatures far above the long - range ordered phases .
this shows the importance of the orbital level structure when analyzing the metal insulator transition in , for example , v@xmath44o@xmath45 .
rp thanks the japan society for the promotion of science ( jsps ) together with the alexander von humboldt - foundation for a postdoctoral fellowship .
this work was supported by kakenhi ( nos .
21740232 , no .
20104010 ) , the grant - in - aid for the global coe programs `` the next generation of physics , spun from universality and emergence '' from mext of japan , and the funding program for world - leading innovative r@xmath82d on science and technology ( first program ) .
tp acknowledges support by the german science foundation ( dfg ) through sfb 602 .
part of the calculations were performed at norddeutsche verbund fr hoch- und hchstleistungsrechnen ( hlrn ) .
35ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop _
( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) http://dx.doi.org/10.1023/a:1003724522132 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) http://stacks.iop.org/0034-4885/38/i=1/a=001 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) `` , '' ( , ) chap .
@noop * * , ( ) | we study the effects of temperature and magnetic field on a two - orbital hubbard model within dynamical mean field theory .
we focus on the quarter filled system , which is a special point in the phase diagram due to orbital degeneracy . at this particular filling the model exhibits two different long - range order mechanisms , namely orbital order and ferromagnetism
. both can cooperate but do not rely on each other s presence , creating a rich phase diagram .
particularly , in the vicinity of the phase transition to an orbitally ordered ferromagnetic state , we observe a strong magnetoresistance effect .
besides the low temperature phase transitions , we also observe a crossover between a paramagnetic insulating and a paramagnetic metallic state for increasing hund s coupling at high temperatures . |
sgr a * , the supermassive black hole at the center of our galaxy , has been observed for several decades .
monitoring stars orbiting around sgr a * has led to measurements of its mass and distance ( ghez et al . 2008 ; gillessen et al . 2009 ) . however , these measurements of mass and distance are strongly correlated . for purely astrometric measurements , mass and distance
are related as @xmath3 , while for measurements of radial velocities mass and distance are related as @xmath4 . for combined data sets , the correlation between mass and distance behaves roughly as @xmath2 ( ghez et al .
2008 ; gillessen et al .
this correlation between mass and distance constitutes a major source of uncertainty in our understanding of the properties of sgr a*. likewise , dynamical measurements of the masses of a number of nearby supermassive black holes have been obtained with often much greater uncertainties ( see , e.g. , gltekin et al .
2009 ) .
another technique , vlbi , aims to image sgr a * directly .
recent vlbi observations with an array consisting of the submillimeter telescope observatory ( smto ) in arizona , the james clerk maxwell telescope ( jcmt ) on mauna kea , and several of the dishes of the combined array for research in millimeter - wave astronomy ( carma ) in california resolved sgr a * on scales comparable to its event horizon and identified sub - horizon size structures ( doeleman et al .
2008 ; fish et al .
images of accretion flows around black holes have the shadow of the compact object imprinted on them , which depends uniquely on its mass , spin , and inclination ( e.g. , falcke et al . 2000 ) as well as on possible deviations from the kerr metric ( johannsen & psaltis 2010 ) . based on such images and assuming the mass and distance obtained from the monitoring of stellar orbits , these vlbi observations inferred constraints on the inclination and spin of sgr a * ( broderick et al .
2009 , 2011 ) and placed limits on potential non - kerr signatures ( broderick et al . 2012 ) .
in addition to the shadow , images of optically thin accretion flows around black holes carry a characteristic signature in the form of a bright ring ( johannsen & psaltis 2010 ) , which we refer to as the photon ring .
light rays that approach the event horizon closely orbit around the black hole many times before they are detected by a distant observer , resulting in a bright ring due to their long optical path length through the accretion flow .
the flux of such photons can account for a significant fraction of the total disk flux and produce higher order images ( cunningham 1976 ; laor , netzer , & piran 1990 ; viergutz 1993 ; bao , hadrava , & stgaard 1994 ; @xmath5 , fanton , & calvani 1998 ; agol & krolik 2000 ; beckwith & done 2005 ) .
these photon rings are clearly visible in all time - dependent general - relativistic simulations of accretion flows that have been reported to date ( mo@xmath6cibrodzka et al . 2009 ; dexter , agol , & fragile 2009 ; shcherbakov & penna 2010 ) .
johannsen & psaltis ( 2010 ) showed that a measurement of the ring diameter measures the ratio @xmath7 for the black hole , independent of its spin or deviation from the kerr metric .
therefore , combining such a measurement with the observations of stars around sgr a * can reduce the correlation between mass and distance . in this paper
, we explore the ability of this approach to refine the mass and distance measurements of sgr a*. we estimate the precision with which a thermal noise - limited vlbi array can infer the diameter of the ring of sgr a * and use a bayesian technique to simulate measurements of the mass and diameter of sgr a * in conjunction with parameters inferred from the existing data of the orbits of stars at comparable wavelengths .
we show that , in this best - case scenario , the correlation between mass and distance is reduced significantly .
in addition , we argue that the accretion flows of other nearby supermassive black holes are optically thin , allowing for vlbi observations of their respective photon rings .
we assess the prospects of using this technique to infer the masses of these sources .
the properties of photon rings are practically independent of the specific flow geometry and remain constant even if the accretion flow itself is highly variable ( johannsen & psaltis 2010 ) . the relative brightness as well as the constancy of these rings make them ideal targets for vlbi - imaging observations . for a kerr black hole with mass @xmath0 , the shape of a given photon ring has a diameter of @xmath8 which remains practically constant for all values of the spin and disk inclination ( johannsen & psaltis 2010 ) . in this expression , @xmath9 is the gravitational radius , and @xmath10 and @xmath11 are the gravitational constant and the speed of light , respectively .
the angular diameter @xmath12 of the diameter of the photon ring of a black hole is given by the ratio of its diameter and distance , @xmath13 assuming the current mass and distance measurements of sgr a * , @xmath14 and @xmath15 ( gillessen et al . 2009 ) , the photon ring has an angular diameter of @xmath16 radio interferometers are limited by their intrinsic resolution as well as by interstellar scattering . in order to identify the range of wavelengths within which vlbi measurements of the photon ring of sgr a * are resolution - limited , we compare the blurring effects of interstellar scattering with the resolution of an interferometer . in figure [ f : openingangle ] we plot the minimum size of resolvable structures on the image of sgr a * using the interstellar scattering law of bower et al .
we also estimate ( dashed line ) the resolution of a radio interferometer at a given wavelength @xmath17 by the expression @xmath18 with @xmath19 and a diameter @xmath20 , which is comparable to the baseline length between the jcmt on hawaii and the south pole telescope .
this yields @xmath21 , where @xmath22 is the resolution in @xmath23arcsec and @xmath24 is the observed wavelength in mm .
dotted lines mark the angular diameters corresponding to length scales of @xmath25 and @xmath26 at the assumed distance of sgr a*. as can be seen from this figure , at sub - mm wavelengths interstellar scattering becomes negligible and measurements are limited by the resolution .
therefore , a measurement of the diameter of the photon ring of sgr a * will require vlbi observations at wavelengths @xmath27 . in the following , we estimate the observed flux of the photon ring of sgr a * assuming a schwarzschild black hole and employing a model of a geometrically thin advection - dominated accretion flow ( adaf ; narayan & yi 1994 , 1995 ; narayan , yi , & mahadevan 1995 ) .
adafs model the accreting gas around a supermassive black hole as a quasi - spherically symmetric plasma consisting of thermal electrons and ions at different temperatures ( narayan & yi 1995a , b ) as well as of non - thermal electrons ( mahadevan 1999 ; zel , psaltis , & narayan 2000 ; yuan , quataert , & narayan 2003 ) .
such an accretion flow is allowed to cool through comptonization and the emission of bremsstrahlung and synchrotron radiation , with the latter generating the predominant contribution to the observed spectrum of sgr a * at radio and wavelengths ( narayan et al .
1995 ) . for our estimate ,
we follow broderick et al .
( 2009 ) and assume an adaf model with a density of thermal electrons @xmath28 electron temperature @xmath29 and magnetic field @xmath30 where @xmath31 . for the coefficients we use ( broderick et al .
2011 ) @xmath32 and @xmath33 these coefficients lead to predictions of the spectrum , polarization , and image size for sgr a * that are in agreement with all current observations .
we assume that all the emission at mm - wavelengths is due to thermal synchrotron radiation ( see narayan et al .
while there may be small contributions of synchrotron emission from non - thermal electrons ( e.g. , mahadevan 1998 ; zel et al .
2000 ; yuan et al .
2003 ) or of other types of radiation at these wavelengths ( such as jets ; e.g. , falcke et al .
1993 ) , this assumption affects our analysis only marginally . in practice
, the measured total flux incorporates all such contributions .
following dolence et al .
( 2009 ) , we write the synchrotron emissivity as @xmath34 where @xmath35 @xmath36 @xmath37 and @xmath38 is the modified bessel function of the second kind of the second order . here , @xmath39 is the angle between the wave vector of the emitted photon and the magnetic field . for our estimate
, we use the average @xmath40 .
we assume that optical paths follow at least one loop along the circular photon orbit of a schwarzschild black hole located at radius @xmath41 .
the emitted intensity of radiation is then given by the expression @xmath42 the observed intensity is related to the emitted intensity by the third power of the redshift factor , which we take to be the gravitational redshift of a photon observed at infinity emitted from @xmath41 ( i.e. , we neglect for this simple estimate the high velocity of the flow , which will serve to increase the intensity at infinity ) .
this yields : @xmath43 where @xmath44 the photon ring has an approximate width of @xmath45 in the image plane ( johannsen & psaltis 2010 ) .
then , the subtended solid angle is given by the expression @xmath46 this yields our estimate for the observed flux density of the photon ring : @xmath47 in figure [ ringflux ] , we plot the modeled ring flux density as a function of wavelength .
we also plot a blackbody function evaluated at the same emission radius and temperature given by expression ( [ temperature ] ) as an upper flux density limit .
the above estimate exceeds the blackbody flux density at wavelengths @xmath48 , and , thus , at these longer wavelengths , the emission becomes self - absorbed .
therefore , we use the minimum of these two flux densities as an estimate of the ring flux density . at wavelengths
@xmath49 , the modeled flux density of the photon ring of sgr a * is @xmath50 , about 1/15 of the total source flux density . since @xmath49 is also in the regime where vlbi observations are resolution - limited ( see figure [ f : openingangle ] ) , this range of wavelengths is optimal for such measurements .
the ring diameter is determined solely by the mass via equation ( [ ringdiameter ] ) , and we can relate the black - hole mass and the angular diameter of the ring according to the expression @xmath51 therefore , for vlbi imaging , the mass is proportional to the distance , and we can reduce the correlation between mass and distance in combination with dynamical measurements .
we now explore the prospect of combining dynamical measurements of sgr a * with vlbi imaging observations of the photon ring .
we analyze the best - case scenario of a thermal noise - limited vlbi array in order to assess whether such a measurement is worthwhile .
systematic limitations will degrade the vlbi measurement somewhat , as discussed below .
we employ a bayesian analysis to estimate the probability distribution over the mass and distance of sgr a * from a measurement of the angular diameter of the photon ring in combination with the constraints obtained from stellar dynamics .
we take the latter as our prior , @xmath52 , by converting into a likelihood the @xmath53 distribution with @xmath54 degrees of freedom obtained from the existing data set of the ephemerides of several s - stars ( gillessen et al . 2009 ) .
we assume a gaussian posterior likelihood of obtaining a particular measurement of the angular diameter as @xmath55,\ ] ] where @xmath12 is given by equation ( [ openingangle ] ) and @xmath56 .
we then use bayes theorem to write the likelihood of a particular mass and distance of the black hole given the data as @xmath57 where @xmath58 is the appropriate constant that normalizes the likelihood . the measurement uncertainty @xmath59 is the key parameter of the likelihood @xmath60 .
in the following , we estimate the scaling of the measurement uncertainty with resolution .
the spatial frequencies of interest are those beyond the first null of the bessel function describing the fourier transform of the ring structure in visibility space . for a given resolution and ring
size , there are @xmath61 accessible half periods of the oscillation , where @xmath61 is given by @xmath62 the typical amplitude of a bessel function oscillation of the first few maxima is of order 0.3 of the peak .
therefore , the signal we seek to measure has an amplitude of @xmath63 where @xmath64 is the observed flux density of sgr a * near 1 mm ( see , e.g. , broderick et al .
2009 ) .
we estimate the expected signal to noise ratio on this measurement by scaling it from the current vlbi measurements .
fish et al .
( 2011 ) measure the size of sgr a * with a signal to noise ratio of @xmath65 .
near - future vlbi arrays will incorporate 5 or 6 stations .
one of them , alma , will have the sensitivity of 50 of the current stations resulting in an overall increase in array sensitivity of a factor of up to roughly 9.3 .
in addition , the scheduled increase of recording bandwidth will increase the sensitivity by a factor of @xmath66 . to account for the variation in system temperature for typical observing conditions and receiver performance
we introduce a degradation in snr proportional to the observing wavelength squared , normalized to the 1.3 mm performance in fish et al .
the total signal to noise ratio of this measurement will then be @xmath67 with the width of the distribution given by the expression @xmath68 in figure [ contours ] , we plot confidence contours of the joint probability distribution for the combination of future thermal noise - limited vlbi and current astrometric observations at three different wavelengths .
the solid line marks the 95% confidence contour determined by the stellar ephemerides ( gillessen et al .
a vlbi measurement at a wavelength of @xmath69 significantly improves the result from stellar orbits alone . at smaller wavelengths , the constraints on the mass and distance of sgr a *
are similar . in the rightmost panel of figure 3
, we have extrapolated the distribution width @xmath59 given by expression ( [ distwidth ] ) to a nominal wavelength of @xmath70 .
measurements at such short wavelengths will be limited by weather conditions and may have to rely on a smaller array with fewer telescopes .
real observations will face more stringent limitations than those imposed by the interferometer thermal noise due to the complications of astrophysics and measurement systematics .
the chief astrophysical limitation is the separation of the ring emission from the source structure in the uv - plane . in our estimate , we have used the location of the nulls as a benchmark for the uncertainties we expect from the vlbi measurement . in practice , however , the full visibility function has to be analyzed with a pattern matching technique that identifies the structure of the ring .
such a technique has to extract the ring from a uv - plane that is only partially sampled by a given set of baselines .
the physics of the accretion flow will also complicate things , as the structure of sgr a * may vary over the course of an observation .
however , because the ring structure is persistent and only weakly altered by rapid changes in the accretion flow we expect that temporal averaging of the visibilities across multiple observing epochs will diminish the importance of such changes .
the vlbi measurement itself must surmount systematic limitations to make the moderate dynamic range measurements proposed here .
chief among these is the difficulty of calibrating the noise level at individual stations , which imposed a 5% uncertainty in fish et al .
( 2011 ) with the three - station array . in observations with the larger array considered here
, there will be many more internal cross - checks available to improve the relative calibration of stations ( the absolute calibration is not important ) .
in particular , the use of three phased interferometers ( mauna kea , carma , alma ) that simultaneously record conventional interferometric data will permit scan - by - scan cross calibration of the amplitude scale of the array . furthermore , the larger arrays will be able to make use of closure phases and closure amplitudes that are immune to calibration errors as part of the ring detection , although we have ignored such procedures here because of the difficulty of simply parameterizing the improvement they can permit .
other effects , such as the coherence of the reference systems between stations ( reported as @xmath71% in fish et al .
2011 ) can be more carefully measured and corrected to prevent them from imposing fundamental limitations to the ring detection .
besides sgr a * , there exist other nearby supermassive black holes , for which a combination of dynamical measurements and vlbi observations could be feasible . since these supermassive black holes are located in host galaxies other than the milky way , observations are much less affected by interstellar scattering . as an example
, broderick & loeb ( 2009 ) and takahashi & mineshige ( 2011 ) analyzed the prospects of imaging the shadow of m87 with vlbi observations at several different wavelengths . in figure
[ smbhs ] , we plot the angular diameter of the photon rings against the distances of a collection of nearby supermassive black holes .
sgr a * is closest to us and has the largest angular diameter , closely followed by m87 and m31 due to their large black hole masses .
the top dashed line indicates the resolution of a telescope array with a baseline equal to the diameter of the earth ( from equation ( [ resolution ] ) ) at a wavelength of 1 mm . for comparison
, we also show the resolution of a future space telescope located at 30% the distance to the moon ( comparable to the orbit of _ chandra _ ) at a wavelength of 0.5 mm corresponding to an angular diameter of about @xmath72 as well as of the proposed _ millimetron _ mission ( wild et al . 2009 ) at a distance of @xmath73 and a wavelength of @xmath74 .
lcccccccc sgr a * & 53 & 0.008 & @xmath75 & 32.48 & 2.4 & 1 & ... & @xmath76 + ngc 4486 ( m87 ) & 22 & 17.0 & @xmath77 & 39.83 & 0.897 & 2 & -5.15 & @xmath76 + ngc 0224 ( m31 ) & 19 & 0.8 & @xmath78 & 32.14 & @xmath79 & 3 & -8.90 & + ngc 4649 ( m60 ) & 13 & 16.5 & @xmath80 & 37.45 & @xmath81 & 4 & -7.84 & @xmath82 + ngc 3115 & 9.6 & 10.2 & @xmath83 & ... & ... & & -7.03 & + ic 1459 & 9.2 & 30.9 & @xmath84 & 39.76 & 0.264 & 5 & ... & + ngc 4374 ( m84 ) & 9.1 & 17.0 & @xmath85 & 38.77 & 0.129 & 6 & -6.29 & + ngc
5128 ( cen a ) & 7.0 & 4.4 & @xmath86 & 39.85 & 6.9 & 7 & ... & + ngc 4594 ( m104 ) & 5.7 & 10.3 & @xmath87 & 37.89 & 0.25 & 8 & -4.68 & + ic 4296 & 2.5 & 54.4 & @xmath88 & 38.59 & 0.155 & 9 & ... & + ngc 1399 & 2.5 & 21.1 & @xmath89 & ... & @xmath90 & 3 & ... & @xmath82 + ngc 4342 & 2.1 & 18.0 & @xmath91 & ... & ... & & ... & + ngc 3031 ( m81 ) & 2.0 & 4.1 & @xmath92 & 36.97 & 0.1812 & 10 & -5.29 & @xmath93 + ngc 4261 & 1.7 & 33.4 & @xmath94 & 39.32 & 0.059 & 11 & -5.21 & + ngc 3585 & 1.6 & 21.2 & @xmath95 & ... & ... & & ... & + ngc 3998 & 1.6 & 14.9 & @xmath96 & 38.03 & @xmath97 & 8 & -4.43 + ngc 4697 & 1.6 & 12.4 & @xmath98 & ... & @xmath97 & 8 & ... & + ngc 4026 & 1.4 & 15.6 & @xmath99 & ... & ... & & ... & + ngc 3379 ( m105 ) & 1.1 & 11.7 & @xmath100 & 35.81 & ... & & -7.57 & + ngc 3245 & 1.0 & 22.1 & @xmath101 & 36.98 & ... & & -5.83 & + ngc 5845 & 1.0 & 28.7 & @xmath102 & ... & ... & & ... & + ngc 3377 & 1.0 & 11.7 & @xmath103 & ... & ... & & -6.16 & + in order to be able to resolve the photon ring with vlbi , the key question is whether the accretion flow of the target supermassive black hole is optically thin . in some cases ,
the spectra of these sources peak at wavelengths near @xmath104 , similar to the spectrum of sgr a * , suggesting that the emission comes from an adaf ( di matteo et al .
2000 ; doi et al . 2005 ) .
naively , we would expect an approximately linear scaling of the electron density of an adaf with the ratio @xmath105 of the black hole , where @xmath106 is its mass accretion rate .
the details of such a relation depend on a variety of factors , such as the temperature profile , the emissivity , and the radiative efficiency
. however , most of the nearby supermassive black holes have very low radiative efficiencies ( ho 2009 ) .
therefore , it is plausible that the accretion flows of these nearby supermassive black holes become optically thin at wavelengths that are comparable to 1 mm making them accessible to vlbi observations .
such observations are best carried out at wavelengths near the flux peak , where the accretion flow is becoming optically thin . at wavelengths
@xmath107 , the emitted flux is likely to be too low to be detected with a vlbi array , while at wavelengths @xmath108 , the accretion flow is optically thick . in table 1 , we summarize the angular diameters , distances , masses , radio luminosities @xmath109 , flux densities @xmath110 near 1 mm , ratios of the bolometric luminosity to the eddington luminosity @xmath111 , and peak wavelengths @xmath112 for supermassive black holes , whose photon rings have an angular diameter of at least @xmath72 .
in addition to sgr a * , the black holes in the centers of m87 , m31 , and m60 are good potential targets for vlbi observations , because of the large angular diameters of their respective photon rings and , in the case of m87 and m60 , the measured peak in the synchrotron part of their spectra near 1 mm .
m87 has a high measured flux density at 86 ghz ( lee et al .
2008 ) and should be readily observable at wavelengths close to 1 mm . in the case of m60 , however , di matteo et al .
( 1999 ) report an upper limit on the flux density of @xmath113 .
no similar flux density measurement of m31 has been reported to date .
we estimate the flux density at 1 mm of m31 from a simple power law fit of from the nasa extragalactic database .
this flux density is relatively small , and m31 as well as m60 may be too faint to be observable with vlbi .
other sources , such as centaurus a , are luminous enough to be detectable at wavelengths near 1 mm . with increasing vlbi resolution , even their photon rings may become observable . in the following ,
we assess the improvement on the mass measurements of the two supermassive black holes ( sgr a * and m87 ) whose photon rings have the largest angular diameters .
we assume fixed distances of @xmath15 ( gillessen et al .
2009 ) and @xmath114 ( gltekin et al . 2009 ) for sgr a * and m87 , respectively . for sgr a * , we estimate an error of the combined mass measurement from the existing stellar ephemerides and our simulated vlbi data of only @xmath115 ( see figure [ contours ] ) . for m87
, we estimate the smallest relative error that thermal noise - limited vlbi imaging observations of the ring can achieve from the signal to noise ratio for observations of sgr a * given by equation ( [ snr ] ) , which we scale with the angular diameter of the black hole to obtain the expression @xmath116 note , however , that m87 has a much larger mass and that the dynamical timescales of its accretion flow is much longer
. therefore , vlbi imaging observations of its photon ring will be much less affected by the variability of the accretion flow as in the case of sgr a*. in table 2 , we compare the relative errors of the mass measurements of sgr a * and m87 ( gillessen et al . 2009 ; gltekin et al . 2009 ) with our estimate of the error of vlbi observations of the respective photon rings at several different wavelengths . in the case of sgr a * , imaging its photon ring improves the error by a factor of about two . in the case of m87 ,
imaging the photon ring at a wavelength of @xmath117 would lead to a result that is similar to the current mass measurement .
lcccc + source & @xmath118 & @xmath119 & @xmath119 & @xmath119 + & & ( 1.0 mm ) & ( 0.8 mm ) & ( 0.5 mm ) + sgr a * & 0.09 & 0.05 & 0.05 & 0.05 + m87 & 0.29 & 1.75 & 0.42 & 0.30 + as we pointed out in section 2 for the case of sgr a * , these errors require further refinement by in - depth imaging simulations . in addition , the morphology of vlbi emission can be complicated by the presence of a jet ( for m87 , see broderick & loeb 2009 ; dexter et al .
2011 ) . as in the case of sgr
a * , the combination of the results from both the dynamical and vlbi imaging observations of m87 would further reduce the error in the masses .
the relative errors of mass measurements of both techniques likewise depend on the error in the measured distances to these sources .
these errors , in turn , depend on uncertainties in the hubble constant , peculiar motions of the gas in host galaxies , as well as assumptions on the proper motion of the milky way ( see , e.g. , hodge 1981 ; jacoby et al .
the details of these effects on the relative errors of the mass are beyond the scope of our analysis .
in this paper , we investigated the prospects of measuring the mass and distance of sgr a * as well as the masses of several other nearby supermassive black holes with a combination of dynamical observations and vlbi imaging of the respective photon rings of these sources . in order to resolve the photon ring of a black hole , its accretion flow must be optically thin .
we argued that the wavelengths at which the accretion flows of these sources become optically thin should be roughly comparable to the location of the peak in the synchrotron emission of sgr a * and identified several supermassive black holes as optimal targets .
we explored the prospects of imaging the photon ring of sgr a * as well as of other nearby supermassive black holes with near - future vlbi arrays .
we estimated the signal to noise ratio with which such arrays can image the photon ring in the best - case scenario if the vlbi observations are limited by thermal noise . based on our estimate , we simulated confidence contours of a mass measurement of sgr a * using existing data of stellar ephemerides .
we showed that the combination of both techniques can indeed reduce the correlation between mass and distance significantly resulting in relative errors of the mass and distance of only a few percent .
we also identified several sources of uncertainty that have to be taken into account for an actual detection of the photon ring of sgr a*. the uncertainties of measurements based on stellar orbits will be further reduced by the continued monitoring and by the expected improvement in astrometry possible with the second generation instrument _ gravity _ for the _ very large telescope interferometer _ ( eisenhauer et al .
further improvements of the vlbi sensitivity will be achieved by the _ event horizon telescope _ , a planned global array of telescopes ( doeleman et al .
2009a , b ; fish et al .
2009 ) .
we estimated the improvement of the mass measurement of m87 using vlbi techniques .
such observations are promising at wavelengths near @xmath117 because of the large size of its photon ring .
for m31 and m60 , the supermassive black holes with the largest photon rings besides sgr a * and m87 , the flux densities may be too low to be detectable with vlbi .
as the resolution of vlbi arrays increases , additional sources will become observable .
angular resolution of @xmath120arcsec requires longer baselines and/or shorter observing wavelengths .
however , the atmosphere precludes regular vlbi observations at wavelengths shorter than @xmath121 at even the best sites .
a measurement of the photon - ring diameter is likely to yield useful results only if the observations extend well beyond the first null in the ring s visibility function . even at a wavelength of @xmath122
, space - vlbi observations will be required to reach this point for all except the first four entries in table 1 .
vlbi between earthbound antennas and a satellite has been achieved at 6 cm wavelength using the japanese _ halca _ satellite ( hirabayashi et al . 1998 ) .
the recently launched _
radioastron _ ( kardashev 2009 ) will extend such observations to 1.2 cm wavelength and baselines as large as @xmath123 , for a resolution of @xmath124 @xmath23arcsec .
a future explorer - class space mission designed to observe at 1 mm wavelength or shorter , where source opacity and scattering effects will be far less important , could provide the angular resolution needed to study a far larger sample of sources .
such a capability may be provided by the russian - european _ millimetron _ mission , which plans to deploy a 12 m antenna with vlbi capabilities to 0.4 mm and maximum baseline @xmath125 km ( wild et al .
2009 ) .
we thank d. zaritzky for useful comments .
tj and dp were supported by the nsf career award nsf 0746549 .
this work was supported at mit haystack observatory by nsf grants ast@xmath126 , ast@xmath127 , and ast@xmath128 .
agol , e. , & krolik , j. h. 2000 , apj 528 , 161 bao , g. , hadrava , p. , & stgaard , e. 1994 , apj , 435 , 55 beckwith , k. , & done , c. 2005 , mnras 359 , 1217 bendo , g. j. , et al .
2006 , apj , 645 , 134 bower , g. c. , goss , w. m. , falcke , h. , backer , d. c. , & lithwick , y. 2006 , apj , 648 , l127 broderick , a. e. , johannsen , t. , loeb , a. , & psaltis , d. , 2012 , in preparation broderick , a. e. , fish , v. l. , doeleman , s. s. , & loeb , a. 2011 , , 735 , 110 broderick , a. e. , fish , v. l. , doeleman , s. s. , & loeb , a. 2009 , , 697 , 45 broderick , a. e. , & loeb , a. 2009 , , 697 , 1164 @xmath5 , a. , fanton , c. , & calvani , m. 1998 , new astronomy , 3 , 647 cunningham , c. 1976 , apj , 208 , 534 dexter , j. , agol , e. , & fragile , p. c. 2009 , apj , 703 , l142 dexter , j. , mckinney , j. c. , & agol , e. 2011 , arxiv:1109.6011 di matteo , t. , fabian , a. c. , carilli , l. c. , & ivison , j. r. 1999 , adv .
space res . , 23 , 1075 di matteo , t. , quataert , e. , allen , s. w. , narayan , r. , & fabian , a. c. 2000 , mnras , 311 , 507 doeleman , s. s. , fish , v. l. , broderick , a. e. , loeb , a. , & rogers , a. e. e. 2009a , apj , 695 , 59 doeleman , s. s. , et al
. 2009b , astro2010 : the astronomy and astrophysics decadal survey , science white papers , no .
68 doeleman , s. s. et al .
2008 , nature , 455 , 78 doi , a. , kameno , s. , kohno , k. , nakanishi , k. , & inoue , m. 2005 , mnras , 363 , 692 dolence , j. c. , gammie , c. f. , mo@xmath129cibrodzka , m. , & leung , p. k. 2009 , apjs , 184 , 387 eisenhauer , f. , et al .
2011 , msngr .
, 143 , 16 falcke , h. , melia , f. , & agol , e. 2000 , apj 528 , l13 falcke , h. , mannheim , k. , & biermann , p. l. 1993 , a&a , 278 , l1 fish , v. l. , et al .
2011 , apj , 727 , 36 fish , v. l. , broderick , a. e. , doeleman , s. s. , & loeb , a. 2009 , apj , 692 , l14 ghez , a. m. et al .
2008 , apj , 689 , 1044 gillessen , s. , eisenhauer , f. , trippe , s. , alexander , t. , genzel , r. , martins , f. , & ott , t. 2009 , apj , 692 , 1075 gltekin , k. , cackett , e. m. , miller , j. m. , di matteo , t. , markoff , s. , & richstone , d. o. 2009 , apj , 706 , 404 hirabayashi , h. , et al .
1998 , science , 281 , 1825 ho , l. c. 2009 , apj , 699 , 626 hodge , p. w. 1981 , a&aa , 19 , 357 israel , f. p. , raban , d. , booth , r. s. , & rantakyr , f. t. 2008 , a&a , 483 , 741 jacobi , g. h. , et al . 1992 ,
pasp , 104 , 599 johannsen , t. , & psaltis , d. 2010 , apj , 718 , 446 kardashev , n. s. 2009 , phys .
u. , 52 , 1127 laor , a. , netzer , h. , & piran , t. 1990 , mnras , 242 , 560 lee , s .- s .
, et al . 2008 , apj , 136 , 159 leeuw , l. l. , sansom , a. e. , robson , e. i. , haas , m. , & kuno , n. 2004 , apj , 612 , 837 mahadevan , r. 1998 , , 394 , 651 middelberg , e. , roy , a. l. , walker , r. c. , & falcke , h. 2005 , a&a , 433 , 897 mo@xmath6cibrodzka , m. , gammie , c. f. , dolence , j. c. , shiokawa , h. , & leung , p. k. 2009 , apj , 706 , 497 narayan , r. , & yi , i. 1995 , apj , 444 , 231 narayan , r. , yi , i. , & mahadevan , r. 1995 , nature , 374 , 623 narayan , r. , & yi , i. 1994 , , 428 , l13 zel , f. , psaltis , d. , & narayan , r. 2000 , apj , 541 , 234 pellegrini , s. , venturi , t. , comastri , a. , fabbiano , g. , fiore , f. , vignali , c. , morganti , r. , & trinchieri , g. 2003 , apj , 585 , 677 sadler , e. m. , et al .
2008 , mnras , 385 , 1656 schdel , r. , krips , m. , markoff , s. , neri , r. , & eckart , a. 2007 , a&a , 463 , 551 shcherbakov , r. v. , & penna , r. f. 2011 , asp conf .
ser . , 439 , 372 takahashi , r. , & mineshige , s. 2011 , apj , 729 , 86 viergutz , s. u. 1993 , astron .
, 272 , 355 wild , w. , et al .
2009 , ex .
a. , 23 , 221 yuan , f. , quataert , e. , & narayan , r. 2003 , , 598 , 301 | dynamical mass measurements to date have allowed determinations of the mass @xmath0 and the distance @xmath1 of a number of nearby supermassive black holes . in the case of sgr
a * , these measurements are limited by a strong correlation between the mass and distance scaling roughly as @xmath2 .
future very - long baseline interferometric ( vlbi ) observations will image a bright and narrow ring surrounding the shadow of a supermassive black hole , if its accretion flow is optically thin . in this paper
, we explore the prospects of reducing the correlation between mass and distance with the combination of dynamical measurements and vlbi imaging of the ring of sgr a*. we estimate the signal to noise ratio of near - future vlbi arrays that consist of five to six stations , and we simulate measurements of the mass and distance of sgr a * using the expected size of the ring image and existing stellar ephemerides .
we demonstrate that , in this best - case scenario , vlbi observations at 1 mm can improve the error on the mass by a factor of about two compared to the results from the monitoring of stellar orbits alone .
we identify the additional sources of uncertainty that such imaging observations have to take into account .
in addition , we calculate the angular diameters of the bright rings of other nearby supermassive black holes and identify the optimal targets besides sgr a * that could be imaged by a ground - based vlbi array or future space - vlbi missions allowing for refined mass measurements . |
the major feature of dynamical processes in many complex relaxing systems is their stochastic background @xcite . particularly , in any dielectric ( complex ) system under an week external electric field ( external action ) only a part ( active dipoles or objects ) of the total number of dipoles is directly governed by changes of the field .
but even those dipoles , not contributing to the relaxation dynamics , can have an effect on the behavior of active dipoles @xcite .
if the dipoles interact with each other , then their evolution has a random character .
consequently , the behavior of such a relaxing system as a whole will not be exponential in nature . in this case
the macroscopic behavior of the complex systems is governed by `` averaging principles '' like the law of large numbers following from the theory of probability @xcite .
the macroscopic dynamics of complex systems is not attributed to any particular object taken from those forming the systems .
the finding out an `` averaged '' object representity for the entire relaxing system is not simple .
the relation between the local random characteristics of complex systems and the universal deterministic empirical laws requires a probabilistic justification .
there are some points of view on this problem .
one of well - known them is based on randomizing the parameters of distributions that describes the relaxation rates in disordered systems . with regard to the dielectric relaxation , each individual dipole in a complex system relaxes exponentially , but their relaxation rates are different and obey a probability distribution ( continuous function ) @xcite .
this approach is successive for getting many empirical response laws and their classification , but it sometime becomes enough complicated to interpret their interrelations and to derive macroscopic response equations .. in this paper we suggest an alternative approach to the analysis of non - exponential relaxation .
it is based on subordination of random processes .
recall that in the theory of anomalous diffusion the notation of subordination occupies one of the most important places ( see , for example , @xcite and references therein ) .
so , a subordinated process @xmath2 is obtained by randomizing the time clock of a random ( parent ) process @xmath3 by means of a random process @xmath4 called the directing process .
the latter process is also often referred to as the randomized time or operational time @xcite . in the common case
the process @xmath5 may be both random and deterministic in nature .
the subordination of random processes is a starting point for the anomalous diffusion theory .
we develop this approach to relaxation processes .
it gives an efficient method for calculating the dynamical evaluating averages of the relaxation processes . in this connection section [ par2 ]
is devoted a presentation of recent achievements of this method .
starting with the description of the two - state system evolution as a markovian process , we develop the analysis on subordinated random processes .
the processes differ from the markovian ones by the temporal variable becoming random . in this context
the cole - cole relaxation is an evident example . in section [ par3 ]
we consider the tempered @xmath0-stable processes .
they overcome the infinite - moment difficulty of the usual ( not tempered ) @xmath0-stable processes .
as applied to the anomalous diffusion , the tempering gives a preserving the subdiffusive behavior for short times whereas for long times the diffusion is something like normal . using the processes in section [ par4 ] , we develop a subordination scheme for the description of the tempered relaxation .
section [ par5 ] formulates major properties of such relaxation .
we show that it has a direct relation to the well - known experimental laws of relaxation , in particular , to the cole - davidson law .
finally , the conclusions are drawn in section [ par6 ] .
the simplest ordinary interpretation of relaxation phenomena is based on the concept of a system of independent exponentially relaxing objects ( for example , dipoles ) with different ( independent ) relaxation rates @xcite .
the relaxation process , following this law ( called debye s ) , may be represented by behavior of a two - state system .
let @xmath6 be the common number of dipoles in a dielectric system . if @xmath7 is the number of dipoles in the state @xmath8 , @xmath9 is the number of dipoles in the state @xmath10 so that @xmath11 .
assume that for @xmath12 the system is stated in order so that the states @xmath8 dominate , namely @xmath13 where @xmath14 is the part of dipoles in the state @xmath8 , @xmath15 the part in the state @xmath10 .
denote the transition rate by @xmath16 defined from microscopic properties of the system ( for instance , according to the given hamiltonian of interaction and the fermi s golden rule ) . in the simplest case ( d relaxation )
the kinetic equation takes the form @xmath17 where , as usual , the dotted symbol means the first - order derivative .
the relaxation function for the two - state system is @xmath18 it is easy see that the steady state of the system corresponds to equilibrium with @xmath19 .
clearly , its response has also an exponential character .
however , this happens to be the case of dipoles relaxing irrespective of each other and of their environment .
if the dipoles interact with their environment , and the interaction is complex ( or random ) , their contribution in relaxation already will not result in any exponential delay .
assume that the interaction of dipoles with environment is taken into account with the aid of the temporal subordination .
we will consider the evolution of the number of dipoles in the states @xmath10 and @xmath8 .
this are parent random processes in the sense of subordination .
they may be subordinated by another random process with a probability density , say @xmath20 .
if @xmath21 and @xmath22 , taking from eq.([eqa1 ] ) as probability laws of the parent processes , depend now on a local time @xmath23 , then the resulting @xmath24 and @xmath25 after the subordination is determined by the integral relation @xmath26 if the directing process ( a new time clock or stochastic time arrow @xcite ) is an inverse @xmath0-stable process , its probability density has the following laplace image @xmath27 where @xmath28 denotes the bromwich path .
this probability density has a simple physical interpretation .
it determines the probability to be at the internal time ( or so - called operational time ) @xmath23 on the real ( physical ) time @xmath29 .
the function @xmath30 can be expanded as a taylor series . besides , it has the fox h - function representation @xmath31 where @xmath32 is the ordinary gamma function . in the theory of anomalous diffusion the random process @xmath33
is applied for the subordination of l@xmath34vy ( or gaussian ) random processes @xcite .
the inverse @xmath0-stable process accounts for the amount of time that a walker does not participate in the motion process @xcite .
if the walker participated all time in the motion process , the internal time and the physical ( external ) time would coincide . as was shown in @xcite ,
the stochastic time arrow can be applied to the general kinetic equation .
then the equation describing a two - state system takes the following form @xmath35 where @xmath36 is the @xmath0-order fractional derivative with respect to time .
here we use the caputo derivative @xcite , namely @xmath37 where @xmath38 means the @xmath39-derivative of @xmath40 .
the relaxation function for the two - state system is written as @xmath41 where @xmath42 is the one - parameter mittag - leffler function @xcite .
it is important to notice that the relaxation function corresponds to the cole - cole ( cc ) law . with reference to the theory of subordination
the cc law shows that the dipoles tend to equilibrium via motion alternating with stops so that the temporal intervals between them is random .
the random values are governed by an inverse @xmath0-stable subordinator .
the evolution of @xmath24 and @xmath15 in eq .
( [ eqb1 ] ) can be connected with the mittag - leffler distribution . if @xmath43 denotes the sum of @xmath39 independent random values with the mittag - leffler distribution , then the laplace transform of @xmath44 is @xmath45 , which tends to @xmath46 as @xmath39 tends to infinity . following the arguments of pillai @xcite
, this supports an infinity divisibility of the mittag - leffler distribution . due to the power asymptotic form ( long tail ) the distribution with parameter @xmath0
is attracted to the stable distribution with exponent @xmath0 , @xmath47 .
the property of the mittag - leffler distribution permits one to determine a stochastic process .
the process ( called mittag - leffler s ) arises of subordinating a stable process by a directing ( generalized ) gamma process @xcite . in this case
the relaxation function has the form @xmath48 where @xmath49 , @xmath50 , @xmath51 are constant .
the one - side fourier transformation of the relaxation function gives the havriliak - negami ( hn ) law @xmath52 this result also corresponds to the well - know hn empirical law .
thus , the hn relaxation can be explained from the subordination approach , if the hitting time process of dipole orientations transforms into the mittag - leffler process @xcite .
for that the hitting time process has an appropriate distribution attracted to the stable distribution . the subordination of the latter results just in the mittag - leffler process .
it is interesting to observe that the lvy process subordinated by another lvy one leads again to the lvy process , but with other index @xcite .
unfortunately , the description from the mittag - leffler process gives nothing for the derivation of any macroscopic response equation like eq.([eqc1 ] ) .
the relaxation model based on the inverse @xmath0-stable process starts with the consideration of @xmath0-stable processes having the infinite - moment difficult . to overcome it
, one can develop an approach stated on tempered @xmath0-stable processes .
the tempered @xmath0-stable process @xcite has the laplace image of its distribution in the form @xmath53 when @xmath54 equals to zero , the tempered @xmath0-stable process becomes simply @xmath0-stable .
however , the distribution ( [ eq5 ] ) describes only probabilistic properties in terms of internal time . for subordination
we need the probability distribution of the inverse tempered @xmath0-stable process . if @xmath55 is the p.d.f . of internal time , then the p.d.f . of its inverse @xmath56
can be represented as @xmath57 taking the laplace transform of @xmath56 with respect to @xmath29 , we get @xmath58 } \,.\label{eq6}\ ] ] when @xmath59 , eq .
( [ eq6 ] ) tends to @xmath60 this expression corresponds to the laplace image of an inverse @xmath0-stable p.d.f . describing a directing process in the theory of cole - cole relaxation .
after the inverse laplace transform we have eq.([eqb1 ] ) . in the laplace space the function @xmath61 has a simple form . because of general properties of the laplace transform we can find @xmath56 explicitly , but its representation is enough complicated and expressed through a integral of the wright functions @xcite . for our calculation it will be sufficient to know only the function @xmath61 .
therefore , we will not present any explicit form @xmath56 here .
if in eq.([eqa1 ] ) the value @xmath16 will depend on time as @xmath62 , we come to the description of the kohlrausch - williams - watts ( kww ) relaxation .
although such a equation does not contain any ( for example , micro / mesoscopic and so on ) details about relaxation processes , it is convenient for practical purpose because of its simplicity . when the relaxation follows the cc ,
cd ( cole - davidson ) , hn laws , the equation ( [ eqa1 ] ) is not so simple as in the case of d and kww relaxation .
recall that the cc relaxation and response functions can be expressed in terms of a solution of the fractional differential equation @xcite . with macroscopic equations for the cd and hn responses
the situation becomes more else complicated .
consider the derivation of the macroscopic response equation of tempered relaxation in more details . for a general type of a markovian process
the general kinetic equation is @xmath63 where @xmath64 denotes the transition rate operator ( see details , in @xcite ) .
this equation defines the probability @xmath65 for the system transition from one state into others .
next , we determine a new process governed by an inverse tempered @xmath0-stable process with the laplace image ( [ eq6 ] ) , namely @xmath66 the laplace transform @xmath67 is given by @xmath68 and leads to @xmath69\,\tilde p\,((s+\delta)^\alpha-\delta^\alpha)-p\,(0)\big\}\nonumber\\ & = & [ ( s+\delta)^\alpha-\delta^\alpha]\,\tilde p_\alpha(s)-\frac{(s+\delta)^\alpha-\delta^\alpha}{s}\,p\,(0)\ , .
\label{eq13}\end{aligned}\ ] ] from this it follows @xmath70 where the kernel @xmath71 is written as @xmath72 where @xmath73 is the two - parameter mittag - leffler function @xcite .
this equation covers a number of particular cases known earlier .
for @xmath74 we obtain eq .
( [ eq12 ] ) , and for @xmath75 it becomes fractional .
if a system has discrete states , then the generating function is of the form @xmath76 where @xmath77 takes values @xmath78 for a series to converge . with the help of the generating function
, one can find the moments by taking the derivative with respect to @xmath77 and then setting @xmath79 .
the generating function of the process governed by the stochastic time clock is given by the relation @xmath80 thus , the generating function for a discrete markovian process directed by a subordinated process can be obtained from the appropriate generating function of the parent process by immediate integration .
the relaxation in a two - state system under the inverse tempered @xmath0-stable subordinator gives @xmath81 for @xmath82 ( or @xmath59 ) the kernel @xmath71 takes the power form @xmath83 as a fractional kernel in the integral representation of eq .
( [ eqc1 ] ) .
however , for @xmath84 ( or @xmath85 ) @xmath71 becomes constant and , as a result , eq .
( [ eq15a ] ) transforms into the integral form of the ordinary equations ( [ eqa1 ] ) . from the linearity of these equations it just follows @xmath86 consequently , we obtain @xmath87\tau^\alpha)\,d\tau,\nonumber\\ n_\downarrow(t)&=&w\int_0^te^{-\delta\tau}\,\tau^{\alpha-1}\,e_{\alpha,\alpha } ( [ \delta^\alpha-2w]\tau^\alpha)\,d\tau.\nonumber\end{aligned}\ ] ] the equations have steady states ( @xmath88 ) corresponding to equilibrium in this system . according to @xcite , we know @xmath89 then @xmath19 for any @xmath90 and @xmath91 ( see , for example , fig .
[ fig1 ] ) .
the transition rate @xmath16 is again defined by microscopic properties of the system .
thus the relaxation function takes the form @xmath92\tau^\alpha)\,d\tau.\ ] ] the response function @xmath93 is written as @xmath94t^\alpha)\,.\ ] ] fig .
[ fig3 ] demonstrates how to change the response function with the increase of @xmath54 . for the experimental study it is interesting to get the frequency - domain representation of the latter function .
its real part describes a dispersion of the relaxing medium , and its imaginary part is an absorption .
the values explicitly can be measured in experiments .
the one - side fourier transformation of the response function gives @xmath95 where @xmath96 and @xmath97 are constant .
the parameter @xmath98 is the characteristic frequency of the relaxing system .
it is easy to notice that the expression ( [ eqe16 ] ) for @xmath74 is reduced to the d law , for @xmath75 ( or @xmath99 ) it describes the cc relaxation , and for @xmath100 it does the cd law ( see fig .
[ fig2 ] ) .
response function with different @xmath54 ( @xmath16 = 0.5 , @xmath0 = 0.5).,width=13 ] imaginary term of the frequency - domain relaxation function @xmath101.,width=13 ]
the frequency dependence of the dielectric susceptibility for orientational polarization of dipoles has been the subject of experimental and theoretical studies for many years , but still there is no any generally accepted theory capable of explaining the observed phenomena .
experimentally it is well known that the complex dielectric susceptibility @xmath101 of most dipolar substances demonstrates a peak in the loss component @xmath102 at a characteristic frequency @xmath103 , and it is characterized by high- ( @xmath104 ) and low - frequency ( @xmath105 ) dependencies .
the tempered relaxation shows @xmath106 where @xmath107 , @xmath108 and @xmath109 . for small @xmath110 and any positive @xmath111 it is easy to see that @xmath112 and @xmath113 whereas for large @xmath110 we get @xmath114 and @xmath115
this implies that @xmath116 where @xmath117 , that is in agreement with the experimental results @xcite .
however , for small @xmath110 we come to @xmath118 this means that the energy lost per cycle does not have a constant relationship to the extra energy that can be stored by a static field .
according to such an asymptotic behavior and in fig .
[ fig2 ] it is seen that the tempered relaxation takes an intermediate place between the d , cc and cd types of relaxation .
in this paper we have represented our progress in the subordination analysis of relaxation phenomena in the complex systems .
the approach permits ones to consider many relaxation laws on the unique theoretical base originating from the stochastic nature of relaxation .
the general probabilistic formalism treats the relaxation of the complex systems regardless of the precise nature of local interactions .
following this approach , we have derived the empirical relaxation laws and their macroscopic equations , have characterized their parameters , have connected the parameters with local random characteristics of the relaxation processes , have demonstrated how to make the transition from the microscopic random dynamics in the complex stochastic systems to the macroscopic deterministic description by integro - differential equations .
it should be pointed out that the form of these equations is a direct sign of complexity evolution in such systems .
although we restricted only by the detailed analysis of two - state systems , due to eq.([eq14 ] ) , this consideration can be developed to the study of many - state systems ( as an example , see the analysis of the three - state fractional system in @xcite ) ) .
the tempered relaxation establishes a connection between several types of relaxation ( d , cc and cd ) .
its asymptotic behavior earnestly shows that starting as a non - exponential relaxation , latter it tends to the d law .
moreover , the theory of subordination suggests a clear interpretation of the tempered relaxation . as applied to the dielectric relaxation , the interaction of dipoles with each other and their environment has a confined time of action on the relaxation process near a starting point of relaxation .
latter they behave independently just as this is the case of exponential ( d ) relaxation . to put it in another way ,
the dipoles are strongly connected with each other in the initial stage of relaxation ( and , therefore , their response function obeys a non - exponential decay ) , but the connection is not long - lived , and in an interval of time each dipole evaluates on its own .
as is grateful to the institute of physics and the hugo steinhaus center for pleasant hospitality during his visit in wrocaw university of technology .
jonscher , _ dielectric relaxation in solids _ ( chelsea dielectric press , london , 1983 ) .
jonscher , _ universal relaxation law _ ( chelsea dielectric press , london , 1996 ) .
a. jurlewicz , k. weron , j. stat . phys . * 73 * , 69(1993 ) ; k. weron , m. kotulski , ibid . * 88 * , 69(1997);a .
jurlewicz , k. weron , a. k. jonscher , ieee trans .
dielectrics * ei8 * , 352(2001 ) ; a. jurlewicz , k. weron j. of non - cryst . sol . * 305 * , 112(2002 ) ; a. k. jonscher , a. jurlewicz , k. weron , contemp . phys . * 44 * , 329(2003 ) .
a. jurlewicz , dissertationes math . *
431 * , 1(2005 ) .
a. jurlewicz , k. weron , cell .
. lett . * 4 * , 55(1999 ) ; a. jurlewicz , k. weron , acta phys . polon .
b * 31 * , 1077(2000 ) .
r. metzler , j. klafter , j. phys .
a * 37 * , r161(2004 ) .
w. feller , _ an introduction to probability and its applications _
( wiley , new york , 1996 ) . c.j.f .
b@xmath119ttcher , p. bordewijk , _ theory of electronic polarization _
( elsevier , amsterdam , 1978 ) .
stanislavsky , phys .
e * 67 * , 021111(2003 ) .
meerschaert , h .-
scheffler , j. appl .
probab . * 41 * , 623 ( 2004 ) ; p. becker - kern , m.m .
meerschaert , h .-
scheffler , ann .
probab . * 32*(1b ) , 730(2004 ) .
b. baeumer , d. a. benson , m. m. meerschaert , physica a * 350 * , 245(2005 ) .
stanislavsky , acta phys .
b * 34*(7 ) , 3649(2003 ) .
stanislavsky , theor . and math
. phys . * 138 * , 418(2004 ) .
m. caputo , j. acousti .
* 66*(1 ) , 176(1979 ) ; r. gorenflo , f. mainardi , `` fractional calculus : integral and differential equations of fractional order '' , in : a. carpinteri and f. mainardi ( eds . ) _ fractals and fractional calculus in continuum mechanics _
( springer - verlag , new york , 1997 ) , pp . 223 - 276 . , edited by a. erd@xmath1lyi ( mcgraw - hill , new york , 1955 ) , vol . 3 , chap . 18 .
pillai , ann .
42*(1 ) , 157(1990 ) .
sokolov , phys .
e * 63 * , 056111(2001 ) .
stanislavsky , chaos , solitons and fractals * 34*(1 ) , 51(2007 ) .
saichev , s.g .
utkin , modern problems of statistical physics bf 1 , 1(2002 ) ( in russian ) ; a. piryatinska , a.i .
saichev , and w.a .
woyczynski , physica a * 349 * , 375(2005 ) .
j. rosiski , stoch .
. appl . * 117 * , 677 ( 2007 ) .
r. gorenflo , yu . luchko , and f. mainardi , fractional calc .
. anal . * 2 * , 383(1999 ) .
n. g. van kampen , _
stochastic processes in physics and chemistry _
( north - holland physics publishing , amsterdam , 1984 ) .
i. podlubny , _ fractional differential equations _ ( academic press , san diego , 1999 ) . | the non - exponential relaxation is shown to result from subordination by inverse tempered @xmath0-stable processes .
the main feature of tempered @xmath0-stable processes is a finiteness of their moments , and the class of random processes includes ordinary @xmath0-stable processes as a particular case .
starting with the cole - cole response this subordination approach establishes its direct link with the cole - davidson law .
we derive the relaxation function describing the tempered relaxation .
the meaning of the empirical response function is clarified .
l@xmath1vy - stable process , subordination , non - exponential relaxation |
the inert doublet model @xcite ( idm ) is one of the simplest extensions of the standard model ( sm ) , yet it is also one of the most versatile .
this scenario , in which the usual sm fields are supplemented by a single scalar @xmath7 doublet which does not contribute to electroweak - symmetry breaking ( ewsb ) and couples to the gauge - boson sector but not the fermion sector , has a wealth of potential phenomenological applications .
perhaps the most intriguing of these stems from the recent observation @xcite that the fields of this additional scalar doublet can provide a positive contribution to the oblique @xmath8 parameter @xcite sufficient to render a sm higgs mass of @xmath9 gev consistent with precision data @xcite .
a host of other potential applications for inert doublets exist as well .
these range from explaining the lightness of neutrino masses via a one - loop radiative see - saw mechanism @xcite to the loop - level induction of electroweak - symmetry breaking @xcite to engineering successful grand unification @xcite .
furthermore , the model yields a natural dark matter candidate in the form of the lightest inert particle ( lip ) , whose absolute stability is guaranteed by an unbroken @xmath10 symmetry .
studies of the relic abundance of lip dark matter @xcite , as well as its prospects for indirect detection via neutrino @xcite , cosmic - ray positron and antiproton @xcite , and gamma - ray @xcite signatures , and for direct detection @xcite have also been performed .
since the coupling structure of the fields of the additional scalar doublet in the idm differs from that of typical two - higgs doublet models ( 2hdm ) in the manner discussed above , the collider phenomenology of the idm also differs markedly from that of such 2hdm .
it is therefore worthwhile to investigate the prospects for detecting the additional fields of the idm via their decay signatures at the large hadron collider ( lhc ) . in this work ,
we focus on the dilepton ( plus missing energy ) channel , which turns out to be one of the most auspicious channels in terms of its discovery potential
. some preliminary , parton - level studies of this channel have been conducted @xcite within one particular region of parameter space . here , we present a more comprehensive , detector - level analysis in which we investigate a variety of different benchmark regions motivated by dark - matter studies , etc .
, and assess the prospects for observing a @xmath11 signal at the lhc in each regime .
we note that the results of this analysis , although conducted in the context of the idm , should also be applicable to other extensions of the sm with similarly - modified scalar sectors , as long as the extra scalars in those extensions have similar decay patterns to those in the idm .
we begin in sect .
[ sec : model ] with a recapitulation of the inert doublet model . in sect .
[ sec : constraints ] , we summarize the theoretical and experimental constraints to which the model is subject .
we outline a set of representative benchmark points which correspond to phenomenologically interesting regions of the parameter space in which all of these constraints are satisfied . in sect .
[ sec : cuts ] , we discuss dilepton production in the idm , as well as the sm backgrounds for @xmath11 at the lhc , and we outline the event - selection criteria we use to differentiate signal events from those produced by these backgrounds .
we present our numerical results in sect .
[ sec : results ] , and in sect .
[ sec : conclusion ] , we conclude .
the inert doublet model is an extension of the sm in which the higgs sector comprises two scalar @xmath7 doublets .
the first of these doublets , here denoted @xmath12 , resembles the sm higgs doublet @xmath13 .
it bears full responsibility for electroweak - symmetry breaking and its neutral component acquires a vacuum expectation value ( vev ) equal to that of the sm higgs : @xmath14 gev .
it also couples to quarks and leptons in precisely the way @xmath13 does in the sm .
the second doublet , here denoted @xmath15 , does not contribute to ewsb ( @xmath16 ) , nor does it have yukawa interactions with the sm quarks and leptons .
this coupling structure is enforced by a @xmath10 symmetry under which @xmath17 and all the sm fields transform trivially .
this symmetry , sometimes called matter parity , remains unbroken in the vacuum , since @xmath18=0 .
the most general @xmath19-even scalar potential allowed by this @xmath10 symmetry is @xmath20 .
\label{eq : v}\ ] ] after ewsb is triggered by the vev of @xmath12 , the scalar spectrum of the model comprises the usual sm higgs @xmath21 ( the neutral , @xmath19-even degree of freedom in @xmath12 ) , as well as four additional fields corresponding to the four degrees of freedom in @xmath15 .
these include a pair of charged scalars @xmath22 , a neutral , @xmath19-even scalar @xmath23 , and a neutral , @xmath19-odd scalar @xmath24 .
the masses of these scalars , given in terms of the six free parameters appearing in eq .
( [ eq : v ] ) is fixed by the constraint @xmath25 from ewsb . ]
@xmath26 in eq .
( [ eq : v ] ) , are @xmath27 for our purposes , it will be useful to parametrize the model using the alternative parameter set @xmath28 , where @xmath29 , @xmath30 , and @xmath31 .
this parameterization is particularly useful in that it characterizes the model in terms of physically significant quantities such as particle masses , mass splittings , and @xmath32 : the coefficient which controls the trilinear @xmath33 and quartic @xmath34 couplings .
the assumption of an unbroken @xmath10 matter parity renders at least one of the degrees of freedom in @xmath15 absolutely stable .
if this particle is electrically neutral ( @xmath23 or @xmath24 in the idm ) , it becomes a good weakly - interacting - massive - particle ( wimp ) dark - matter candidate .
for the remainder of this work , we will assume that the lip is the @xmath19-even scalar @xmath23 , and hence @xmath35 .
the phenomenology of the alternative scenario , in which the @xmath19-odd scalar @xmath24 plays the role of the lip , is very similar .
a variety of considerations , stemming both from theoretical consistency conditions and from experimental bounds , constrain the idm .
below , we briefly summarize these constraints , which were discussed in detail in ref .
@xcite . *
* perturbativity : * @xmath36 * * vacuum stability : * @xmath37 * * limits from direct collider searches : * + first of all , the excellent agreement between the experimentally - measured values for @xmath38 and @xmath39 obtained from lep and tevatron data @xcite and the predictions of the sm requires that @xmath40 in order that the decays @xmath41 and @xmath42 are kinematically forbidden .
+ second of all , bounds on the invisible decays of the higgs boson from lep data @xcite also serve to constrain scenarios in which the higgs is light and @xmath43 . in this work , however , we will be primarily concerned with cases in which @xmath44 gev , for which the bounds from the searches on invisible higgs decay does not apply .
+ third and finally , one can consider limits arising from direct searches for @xmath22 , @xmath24 , and @xmath23 , both at lep and at the tevatron @xcite . it should first be noted that the standard limits on additional charged and neutral higgs scalars do not apply , because the standard search channels from which they are derived generally involve the couplings of such scalars to fermions , which are absent in the idm . on the other hand , bounds derived from the non - observation of @xmath45 @xcite and @xmath46
@xcite decays in supersymmetric models can be used to constrain the idm parameter space , since @xmath47 and @xmath48 in the idm lead to similar signals .
a detailed analysis of the constraints on @xmath47 in the idm based on lep ii searches for @xmath45 was conducted in ref .
@xcite , which showed that regions of parameter space with @xmath49 gev and @xmath50 gev for @xmath51 gev had been ruled out . for @xmath52 gev , however , only the lep i constraint @xmath53 applies .
a rough bound of @xmath54 gev @xcite can also be derived from the lep @xmath46 limit by making the necessary modifications to account for the difference in cross - section between fermion - pair and scalar - pair production .
taking these considerations into account , we will henceforth restrict our attention to models for which @xmath55 gev . *
* electroweak precision constraints : * + electroweak precision measurements set limits on contributions from the additional higgs doublet to the oblique @xmath23 and @xmath8 parameters @xcite .
we consider a given parameter choice to be consistent with electroweak precision constraints as long as the overall values of @xmath23 and @xmath8 it yields , once all contributions are taken into account , lie within the 68% c.l .
ellipse determined by the lep electroweak working group @xcite . for a light sm higgs , with @xmath56 gev , the constraint is weak as long as @xmath57 and @xmath58 are of roughly the same order . for a heavy sm higgs ,
a large splitting between @xmath22 and @xmath23 is preferred , and @xmath59 . *
* dark matter relic density : * + one of the attractive aspects of the inert doublet model is that it can provide a viable wimp dark matter candidate in the form of a stable , neutral lip . the model is therefore constrained by experimental limits on the relic density of dark matter in the universe . in
what follows , we will assume that the lip relic density represents the dominant component of @xmath60 and falls within the @xmath3 range of the dark - matter density of the universe as measured by wmap @xcite , only the upper bound applies .
] : @xmath61 .
a detailed examination of the relic density of a @xmath19-even scalar lip in the idm was conducted in @xcite .
it was found that the correct dark - matter relic density could be realized in several distinct regions of parameter space in which all the aforementioned theoretical and experimental constraints were also satisfied . for a light sm higgs with @xmath62 gev ,
two scenarios are possible .
the first of these involves a light lip with @xmath63 gev and mass splittings @xmath64 and @xmath65 which are sizeable , but of the same order .
the second involves a heavier dark matter particle with @xmath66 gev and relatively small mass splittings . for a heavy sm higgs with @xmath67 gev ,
the regions which the constraints leave open are those in which @xmath68 gev and @xmath59 , with both @xmath57 and @xmath58 relatively large , or @xmath69 gev , @xmath52 gev with a large @xmath57 .
.a list of benchmark points used in our analysis , defined in terms of the model parameters @xmath70 .
dark matter relic density and collider phenomenology of the idm depend little on @xmath71 , which is set to 0.1 for all benchmark points
. the points lh1
@xmath72 lh5 involve a light ( @xmath73 ) higgs boson , while the points hh1 @xmath72 hh3 involve a heavy ( @xmath74 ) higgs . [ cols="^,^,^,^,^,^",options="header " , ] let us now turn to examine the results for each of the individual benchmark scenarios in our study in more detail .
we begin with the lh1 , which involves a light higgs boson , a light lip ( @xmath75 gev ) , and a large mass splitting ( @xmath76 gev @xmath77 ) .
the dominant backgrounds in this scenario are those from @xmath78 and @xmath79 , each of which also involves the leptonic decay of an on - shell @xmath80 and is therefore difficult to differentiate from the signal process on the basis of kinematical variables .
the remaining backgrounds are efficiently suppressed after the imposition of the @xmath81 cut near the @xmath6 window . with @xmath82 of integrated luminosity
, a significance level of @xmath83 could be obtained in this benchmark scenario .
the situation for the heavy - higgs benchmark hh1 is similar ; however the smaller @xmath84-production cross - section in this case ( due primarily to an increased lip mass ) translates into a lower statistical significance .
benchmark point lh2 also includes a @xmath0 gev dark matter particle , but involves a smaller mass splitting than that of lh1 : @xmath85 gev .
this scenario affords the best opportunity for discovery at the lhc out of any of the benchmark points in our analysis , yielding a statistical significance of @xmath86 with @xmath5 of integrated luminosity .
two factors contribute to its success : a small production threshold @xmath87 , and the fact that @xmath88 , which implies that the @xmath19-odd scalar @xmath24 decays via an off - shell @xmath80 .
the latter consideration makes it possible to eliminate @xmath78 and @xmath79 background contributions quite efficiently by setting @xmath89 comfortably below the @xmath80 pole .
further cuts on the angular variables @xmath90 and @xmath91 serve to reduce the remaining backgrounds to a manageable level .
after all cuts are imposed , events from the low-@xmath81 tail of the @xmath78 distribution form the dominant background .
it should be noted , however , that while the aforementioned angular - separation cuts are quite efficient in reducing background events , this efficiency comes with a price : the cuts also eliminate a substantial fraction of signal events .
this explains why the signal cross - section for lh2 is less than that for lh1 , as no angular separation cuts are imposed in the latter scenario .
benchmarks lh3 and lh4 are superficially similar , given that they involve a similar lip mass @xmath92 gev and the same mass splitting @xmath93 gev . in this case , however , the marked difference in @xmath57 a parameter which generally has little effect on observability of the dilepton signal in the @xmath84 associated - production channel between the two points has a substantial impact on their collider phenomenology .
the reason for this is twofold .
first of all , since @xmath94 for lh4 ( unlike any other benchmark in our analysis ) , the decay channels @xmath95 , where @xmath96 denotes either four jets , two jets and a single charged lepton , or two charged leptons , are open in this scenario , with a branching ratio @xmath97 . as a result , @xmath98 , and thus the dilepton signal cross - section ,
are reduced by an additional factor of two relative to those points for which such competing decays are kinematically prohibited .
second of all , as discussed above , the small value for @xmath99 gev in lh4 allows the additional contribution of @xmath100 process to the signal due to the unobservable soft jets and leptons from @xmath22 decay .
these additional contributions augment the overall signal cross - section and more than compensate for the diminished @xmath101 , as discussed above . for @xmath5 of integrated luminosity , a significance level greater than @xmath3 could be reached for lh4 as well as lh3 .
the final light - higgs scenario in our analysis , lh5 , turns out to be the most difficult benchmark point for which to observe a dilepton signal , primarily because of the small mass splitting @xmath102 gev between @xmath23 and @xmath24 .
the charged leptons in the final state tend to be extremely soft , and consequently the signal remains buried under the sm background even after an optimized set of level iii cuts is applied .
scenarios with a small value of @xmath58 will in general be difficult to discover via this channel at the lhc .
it should be noted that the results we obtain for this benchmark differ significantly from the parton - level results quoted in @xcite for a similar benchmark scenario , also with @xmath102 gev .
the discrepancy owes primarily to our imposition of a level i cut of @xmath103 cut designed to replicate the effect of electron and muon isolation requirements at the atlas and cms detectors .
since the angular separation between the lepton momenta tends to be extremely small for such a small value of @xmath58 , a vast majority of signal events will have @xmath104 and hence be eliminated by this cut .
let us now turn to discuss the benchmark points which feature a heavy ( @xmath105 gev ) higgs boson in other words , those benchmarks for which the idm successfully addresses the lep paradox .
while the electroweak precision constraints discussed in sect . [ sec : constraints ] are more stringent in this case , these constraints primarily affect @xmath57 , which is typically required to be quite large .
since this parameter generally does not affect results in the dilepton channel , which depend primarily on @xmath106 and @xmath58 , the same qualitative results obtained for the light - higgs benchmarks also apply here . for hh1 , with @xmath76 gev , a significance level of only @xmath107 can be achieved with @xmath5 of integrated luminosity , due to both the overwhelming sm backgrounds that exist for dilepton processes involving on - shell @xmath80 decay , and a suppressed signal cross - section relative to benchmark lh1 ( which has a far lighter lip ) . for hh3 a benchmark with a somewhat small value of @xmath58 a @xmath108 gev cut helps to cut down the sm backgrounds from processes involving on - shell @xmath80 decay .
it is , however , hard to improve upon the statistical significance by implementing additional cuts .
the remaining background events which survive this cut ( most of which come from @xmath109 ) tend to have similar @xmath90 and @xmath91 distributions to those of the signal a situation which makes the application of further , angular cuts essentially redundant .
furthermore , since the missing - energy distribution for the signal events in scenarios with small @xmath58 peaks at relatively low values of @xmath110 , there is little to be gained by increasing @xmath111 much beyond the level ii threshold of @xmath112 gev .
by contrast , in scenarios with larger @xmath58 , an elevated missing - energy cut works quite effectively in tandem with the angular cuts in reducing backgrounds from @xmath109 and @xmath113 .
a significance level of @xmath114 is reached for hh3 with @xmath5 of integrated luminosity .
it is benchmark hh2 , however , which affords the best opportunity for discovery at the lhc from among the heavy - higgs scenarios , with a statistical significance of @xmath115 at @xmath5 of integrated luminosity .
this is because the signal for this benchmark can be distinguished from the @xmath79 and @xmath78 backgrounds on the basis of @xmath81 cuts , and from the remaining @xmath109 , @xmath116 , and @xmath113 backgrounds on the basis of @xmath90 , @xmath91 , and @xmath110 cuts in the same manner as for the low - higgs - mass point lh2 .
we therefore conclude that even scenarios in which the idm permits an evasion of the lep upper bound on @xmath117 can yield an observable dilepton signal at the lhc . from the results in table [ tab : bmxsecscutiii ] ,
it is evident that the prospects for detecting a signal in the dilepton channel in the idm model hinge primarily on two criteria .
the first of these is the dependence of the cross - section for @xmath118 on @xmath87 and @xmath58 .
this cross - section is , of course , larger in cases where the pair - production threshold energy @xmath87 is small . among cases with similar values of @xmath87 ,
those in which @xmath58 is smaller will have larger production cross - sections .
this can be understood by noting that the partonic cross - section for this process depends on @xmath106 and @xmath58 in the following way : @xmath119^{3/2}. \label{eq : partonicxsec}\end{aligned}\ ] ] for values of @xmath120 , for which the dependence of this expression on @xmath106 and @xmath58 is non - negligible , it is apparent that for fixed @xmath121 , @xmath122 decreases with increasing @xmath58 .
this accounts for the difference between the @xmath123 production cross - sections for benchmarks lh1 and hh3 quoted in table [ tab : bmxsecs ] .
the second criterion is the relationship between @xmath58 and @xmath6 : cases in which @xmath88 tend to have a higher statistical significance than those in which @xmath124 , as is manifest from comparing the results for benchmarks lh2 and lh1 in table [ tab : bmxsecscutiii ] .
this is because in the latter case , it is difficult to distinguish the signal process from the dominant @xmath78 background on the basis of event topology . on the other hand , when @xmath58 is exceedingly small ( as it is in our lh5 scenario ) , the charged leptons will be so soft that the detector - acceptance ( i.e. level i ) cuts will eliminate the vast majority of would - be signal events , as discussed above . between these extremes ,
a window of @xmath125 emerges within which the prospects for observing a signal are quite good , so long as the lip mass also falls roughly within the @xmath2 gev range .
for cases in which @xmath126 , the prospects for discovery at the lhc are reasonable meaning a statistical significance around the @xmath3 level with 100 @xmath127 of integrated luminosity only if the dark - matter particle is light ( @xmath75 gev ) .
it is not difficult simultaneously to satisfy the constraints discussed in sect .
[ sec : constraints ] and to realize a @xmath58 value within this mass window of @xmath2 gev while keeping the lip mass relatively light ( @xmath49 gev ) or , alternatively , to obtain a large mass splitting @xmath128 and a light lip mass of around 40 gev .
this is true not only in models where the higgs boson is light and the parameters of the theory comparatively unconstrained , but also in cases in which the mechanism of ref .
@xcite for evading electroweak precision bounds on the higgs mass is realized in nature , and @xmath129 gev . in either case
, it would be possible to observe a dilepton signal at the lhc at a significance level of @xmath3 or higher , with an integrated luminosity of @xmath5 .
in this work , we have investigated the potential for observing a dilepton signature in the inert doublet model at the lhc .
we have explored the prospects for a number of benchmark scenarios , including several in which the idm successfully ameliorates the lep paradox and the higgs - boson mass can be elevated as high as @xmath130 gev , as well as several of the dark - matter motivated scenarios cataloged in ref .
we have shown that for cases in which the dark matter candidate is relatively light ( @xmath2 gev ) and @xmath131 gev , a signal with a significance of more than @xmath3 should be apparent at the lhc with less than @xmath5 of integrated luminosity . moreover , in cases when the lip is on the lighter end of this range , a @xmath3 discovery would be possible with only @xmath132 of integrated luminosity .
in addition , there are also certain cases in which @xmath124 and the lip is light ( @xmath75 gev ) for which the prospects for detection are also reasonably good .
of course the observation of an excess in the @xmath133 channel alone , while exciting , is by no means conclusive evidence for the inert doublet model . indeed , many models of beyond - the - standard - model physics lead to such a signature , including weak - scale supersymmetry , two - higgs - doublet models , etc .
fortunately , evidence for the idm can come from a number of other sources .
some of these sources involve other channels associated with the sm - like higgs at the lhc .
one potentially interesting signal could arise due to deviations of the decay properties of the higgs boson @xmath21 from those of a sm higgs . in situations
in which @xmath134 , for example , @xmath135 can contribute substantially to the invisible higgs width .
searches for the weak - boson fusion ( wbf ) process @xmath136 , with @xmath21 decaying invisibly , can be used effectively to identify a higgs boson at the lhc @xcite , and preliminary studies @xcite indicate that a @xmath137 discovery should be possible with only @xmath132 of integrated luminosity in regions of parameter space where @xmath138 is large . moreover , if @xmath139 , the tetralepton + @xmath140 signatures resulting from decays of the form @xmath141 may also be detectable in certain regions of parameter space .
the observation of signals of this sort , along with the non - observation of other signals which appear in standard 2hdm due to @xmath142 couplings ( where @xmath143 and @xmath144 and @xmath145 are sm fermions ) absent in the inert doublet model , could together serve to distinguish the idm from other scenarios for physics beyond the standard model .
evidence for the idm could also come from a a variety of other sources , including dark - matter - direct - detection experiments and from the observation of energetic gamma - rays @xcite or neutrinos @xcite resulting from lip dark matter annihilation .
clearly , the particular set of signals that an inert doublet would manifest differs substantially , depending on which of the allowed regions of parameter space the model happened to inhabit , and as we have shown , the @xmath11 channel can provide an important probe into which region that might be .
we would like to thank j. alwall and t. han for useful discussions and comments .
we would like to thank the aspen center of physics , where part of this work was completed .
this work was supported in part by the department of energy under grant de - fg02 - 04er-41298 .
n. g. deshpande and e. ma , phys .
d * 18 * , 2574 ( 1978 ) .
r. barbieri , l. j. hall and v. s. rychkov , phys .
d * 74 * , 015007 ( 2006 ) [ arxiv : hep - ph/0603188 ] . m. e. peskin and t. takeuchi , phys .
d * 46 * , 381 ( 1992 ) .
m. w. grunewald , j. phys .
ser . * 110 * , 042008 ( 2008 ) [ arxiv:0709.3744 [ hep - ph ] ] .
e. ma , phys .
d * 73 * , 077301 ( 2006 ) [ arxiv : hep - ph/0601225 ] .
t. hambye and m. h. g. tytgat , phys .
b * 659 * , 651 ( 2008 ) [ arxiv:0707.0633 [ hep - ph ] ] .
m. lisanti and j. g. wacker , arxiv:0704.2816 [ hep - ph ] .
l. lopez honorez , e. nezri , j. f. oliver and m. h. g. tytgat , jcap * 0702 * , 028 ( 2007 ) [ arxiv : hep - ph/0612275 ] .
e. m. dolle and s. su , arxiv:0906.1609 [ hep - ph ] .
p. agrawal , e. m. dolle and c. a. krenke , phys .
d * 79 * , 015015 ( 2009 ) [ arxiv:0811.1798 [ hep - ph ] ] .
e. nezri , m. h. g. tytgat and g. vertongen , jcap * 0904 * , 014 ( 2009 ) [ arxiv:0901.2556 [ hep - ph ] ] . m. gustafsson , e. lundstrom , l. bergstrom and j. edsjo , phys . rev
. lett . * 99 * , 041301 ( 2007 ) [ arxiv : astro - ph/0703512 ] .
e. dolle and s. su , in preparation .
d. majumdar and a. ghosal , mod .
lett . a * 23 * , 2011 ( 2008 ) [ arxiv : hep - ph/0607067 ] . q. h. cao , e. ma and g. rajasekaran , phys . rev .
d * 76 * , 095011 ( 2007 ) [ arxiv:0708.2939 [ hep - ph ] ] .
w. m. yao _ et al .
_ , journal of physics g * 33 * , 1 ( 2006 ) , partial update for 2008 available at http://pdg.lbl.gov/. [ lep higgs working for higgs boson searches and aleph collaboration and delphi collaboration and cern - l3 collaboration and opal collaboration ] , arxiv : hep - ex/0107032 .
[ lep higgs working group for higgs boson searches and aleph collaboration and aleph collaboration and delphi collaboration and l3 collaboration and opal collaboration ] , arxiv : hep - ex/0107031 ; j. abdallah _ et al .
_ [ delphi collaboration ] , eur .
j. c * 34 * , 399 ( 2004 ) [ arxiv : hep - ex/0404012 ] .
a. abulencia _ et al .
_ [ cdf collaboration ] , phys . rev .
lett . * 96 * , 042003 ( 2006 ) [ arxiv : hep - ex/0510065 ] . m. acciarri _ et al . _
[ l3 collaboration ] , phys .
b * 472 * , 420 ( 2000 ) [ arxiv : hep - ex/9910007 ] ; r. barate _ et al . _
[ aleph collaboration ] , phys .
b * 499 * , 67 ( 2001 ) [ arxiv : hep - ex/0011047 ] ; j. abdallah _ et al .
_ [ delphi collaboration ] , eur .
j. c * 31 * , 421 ( 2004 ) [ arxiv : hep - ex/0311019 ] ; g. abbiendi _ et al . _
[ opal collaboration ] , eur .
j. c * 35 * , 1 ( 2004 ) [ arxiv : hep - ex/0401026 ] .
the lep susy working group , http://lepsusy.web.cern.ch/lepsusy/ , lepsusywg/01 - 03.1 .
e. lundstrom , m. gustafsson and j. edsjo , arxiv:0810.3924 [ hep - ph ] .
a. pierce and j. thaler , jhep * 0708 * , 026 ( 2007 ) [ arxiv : hep - ph/0703056 ] .
[ aleph collaboration and delphi collaboration and l3 collaboration and opal collaboration and sld collaboration and lep electroweak working group and sld electroweak group and sld heavy flavour group ] , phys . rept . * 427 * , 257 ( 2006 ) [ arxiv : hep - ex/0509008 ] .
d. n. spergel _ et al . _
[ wmap collaboration ] , astrophys .
j. suppl .
* 148 * , 175 ( 2003 ) [ arxiv : astro - ph/0302209 ] .
r. arnowitt and p. nath , mod .
lett . a * 2 * , 331 ( 1987 ) ; h. baer and x. tata , phys
. rev .
d * 47 * , 2739 ( 1993 ) ; h. baer , c. kao and x. tata phys .
d * 48 * , 5175 ( 1993 ) [ hep - ph/9307347 ] ; t. kamon , j. lopez , p. mcintyre and j.t .
white , phys .
d * 50 * , 5676 ( 1994 ) [ hep - ph/9406248 ] ; h. baer , c .- h .
chen , c. kao and x. tata , phys .
d * 52 * , 1565 ( 1995 ) [ hep - ph/9504234 ] ; s. mrenna , g.l .
kane , g.d .
kribs and j.d .
wells , phys .
d * 53 * , 1168 ( 1996 ) [ hep - ph/9505245 ] .
matchev and d.m .
pierce , phys .
b * 467 * , 225 ( 1999 ) [ hep - ph/9907505 ] ; phys . rev .
d * 60 * , 075004 ( 1999 ) [ hep - ph/9904282 ] .
k. t. matchev and d. m. pierce , phys .
b * 467 * , 225 ( 1999 ) [ arxiv : hep - ph/9907505 ] .
j. alwall _ et al .
_ , jhep * 0709 * , 028 ( 2007 ) [ arxiv:0706.2334 [ hep - ph ] ] . t. sjostrand , s. mrenna and p. skands , jhep * 0605 * , 026 ( 2006 ) [ arxiv : hep - ph/0603175 ] .
`` pgs pretty good simulator '' , + http://www.physics.ucdavis.edu/ @xmath146conway / research / software / pgs / pgs4-general.html [ atlas collaboration ] , atlas - tdr-14 ; [ atlas collaboration ] , atlas - tdr-15 .
g. l. bayatian _ et al . _
[ cms collaboration ] , j. phys .
g * 34 * , 995 ( 2007 ) .
o. j. p. eboli and d. zeppenfeld , phys .
b * 495 * , 147 ( 2000 ) [ arxiv : hep - ph/0009158 ] . | the inert doublet model is one of the simplest and most versatile scenarios for physics beyond the standard model . in this work ,
we examine the prospects for detecting the additional fields of this model at the lhc in the dilepton channel .
we investigate a wide variety of theoretically- and phenomenologically - motivated benchmark scenarios , and show that within regions of model parameter space in which the dark - matter candidate is relatively light ( between @xmath0 and @xmath1 gev ) and the mass splitting between the neutral scalars is also roughly @xmath2 gev , a signal at the @xmath3 to @xmath4 significance level can be observed with @xmath5 of integrated luminosity .
in addition , even if the mass splitting between the neutral scalars is larger than @xmath6 , a signal of more than @xmath3 can be observed as long as the mass of the dark - matter candidate is around 40 gev . .7ex .7ex
= cmss10 = cmss10 at 7pt |
yangian for @xmath4 is the associative algebra historically one of the first examples of _ quantum groups_. the yangian @xmath1 is a hopf algebra deforming the enveloping algebra @xmath5)$ ] , where @xmath6 $ ] is the ( infinite dimensional ) lie algebra of @xmath4-valued polynomials .
this algebra was considered in the works of l. fadeev and st .- petersburg school in the relation with the inverse scattering method , see e.g. @xcite . there is a family @xmath0 of commutative subalgebras in @xmath1 parameterized by complex matrices @xmath7 called bethe subalgebras .
this family originates from the integrable models in statistical mechanics and algebraic bethe ansatz . for details and links on yangians , we refer the reader to the survey @xcite and to the book @xcite by a. molev .
denote by @xmath8 the maximal torus in @xmath9 i.e. the subgroup of diagonal matrices in @xmath9 . in the present paper
we restrict ourselves to bethe subalgebras with @xmath10 .
let @xmath11 be the set of regular elements of the torus , i.e. the set of matrices from @xmath8 with pairwise distinct eigenvalues .
in @xcite nazarov and olshanski showed that @xmath0 is a free polynomial algebra and that it is a maximal commutative subalgebra in @xmath1 for all @xmath12 . for non - regular @xmath13 ,
the subalgebra @xmath0 becomes smaller .
but there is a natural way to obtain commutative subalgebras of the same size as for @xmath12 by taking some _ limits _ of regular ones .
for example , one can get the gelfand - tsetlin subalgebra ( also known as cartan subalgebra ) of the yangian as a limit of some @xmath14-parametic family of bethe subalgebras with @xmath15 .
our goal is to describe all limit subalgebras . the same question ( to describe all limit subalgebras ) for the images of bethe subalgebras in the universal enveloping algebra @xmath16 under the evaluation homomorphism ( under the name of `` shift of argument subalgebras '' )
was asked by vinberg in late 1990-s .
the answer was given by v.shuvalov in @xcite and later in more algebro - geometric terms by l. aguirre , g. felder and a. veselov in @xcite .
their description is roughly as follows .
shift of argument subalgebras themselves are parametrized by regular diagonal @xmath17-matrices up to proportionality and up to adding a scalar matrix .
the latter can be regarded as the space of configurations of @xmath18 pairwise distinct points on the complex line .
it turns out that the limit shift of argument subalgebras are parametrized by the deligne - mumford closure of this space @xcite , all limit subalgebras are free @xcite and maximal commutative @xcite , and moreover there is an explicit inductive procedure generating the limit subalgebras from smaller shift of argument subalgebras assigned to smaller @xmath18 .
it is natural to expect a similar description for limit bethe subalgebras in the yangian .
the construction of bethe algebras can be regarded as a regular map from @xmath11 to the `` grassmannian '' of subspaces in @xmath1 of the same `` dimension '' as @xmath0 .
the space @xmath11 is noncompact while the grassmannian is ( in appropriate sense ) compact .
so we can take the closure of the image of this map and obtain new subalgebras with the same poincare series .
we call such subalgebras _ limit subalgebras .
_ since the subalgebras @xmath0 do not change under dilations of @xmath2 , the parameter space for the family @xmath0 is the quotient @xmath19 of the set of regular elements of the torus by the subgroup of scalar matrices . following @xcite we can regard the space @xmath19 as the moduli space @xmath20 of rational curves with @xmath21 marked points ( we can assign to a matrix @xmath2 with the eigenvalues @xmath22 the curve @xmath23 with the marked points @xmath24 ) .
the limit subalgebras of the family @xmath0 are thus parametrized by some compactification of the space @xmath20 .
the main results of the present paper can be summarized as follows : * main theorem . *
_ the closure of @xmath19 which parametrizes the limit subalgebras is isomorphic to the deligne - mumford compactification @xmath3 .
all limit subalgebras are free polynomial algebras and maximal commutative subalgebras in @xmath1 .
_ in fact we will describe all limit subalgebras as products of smaller bethe subalgebras and some shift of argument subalgebras ( which are the images of bethe subalgebras in the universal enveloping algebra @xmath16 under the evaluation map ) .
the main idea of the proof is to reduce this problem to the similar one for the family of shift of argument subalgebras in @xmath25 where @xmath26 is big enough . for this
we use on the one hand the olshanski centralizer construction @xcite which approximates the yangian @xmath1 by the centralizer subalgebras of the form @xmath27 , and , on the other hand , the results of shuvalov @xcite and tarasov @xcite which describe the limit subalgebras for the family of shift of argument subalgebras .
let @xmath28 be the image of @xmath29 under the evaluation homomorphism @xmath30 . the subalgebra @xmath31 does not change after adding a scalar matrix to @xmath2 .
the associated graded of @xmath31 is a poisson commutative subalgebra @xmath32 known as _ shift of argument subalgebra
_ since it is generated by the adjoint invariants from @xmath33 shifted by @xmath34 for all @xmath35 .
suppose @xmath2 is a diagonal matrix with pairwise distinct eigenvalues @xmath22
. then the algebra @xmath36 contains the quadratic elements @xmath37 which are the coefficients of ( an appropriate version of ) the @xmath38 connection .
moreover both @xmath36 and @xmath31 are uniquely determined by the linear span @xmath39 of the @xmath40 s .
note that @xmath40 do not change under simultaneous affine transformations of the @xmath41 s , hence the space of parameters of the corresponding shift of argument subalgebras is naturally the configuration space of @xmath18 pairwise distinct points on the affine line or , equivalently , the configuration space @xmath42 of @xmath43 pairwise distinct points on the projective line . from the results of aguirre , felder and veselov @xcite it follows that the closure of the family of subspaces @xmath44 is the deligne - mumford compactification @xmath45 , which is crucial for our proof .
let @xmath46 $ ] be the filtered polynomial algebra of infinitely many generators such that @xmath47 .
the olshanski centralizer construction @xcite is the collection of surjective homomorphisms of filtered algebras @xmath48 generalizing the evaluation map .
the intersection of kernels of such homomorphisms is known to be zero , so this collection of homomorphisms can be regarded as an _ an asymptotic isomorphism _
i.e. for each filtered component of @xmath49 there is @xmath50 such that for any @xmath51 the restriction of the above homomorphism to this filtered component is an isomorphism .
the idea of our proof is to analyze the parameter spaces of the images of @xmath0 under the centralizer construction maps . since centralizer construction is an asymptotical isomorphism the closure of the parameter space stabilizes for @xmath52 .
we prove that the image of @xmath53 in the centralizer algebra @xmath27 is contained in some non - regular shift - of - argument subalgebra @xmath54 ( proposition [ incl ] ) and moreover @xmath54 is asymptotically isomorphic to @xmath53 .
we show that the closure of the parameter space for the subalgebras @xmath54 is @xmath3 ( realized as a subvariety in @xmath55 which parametrizes all limit shift of argument subalgebras in @xmath56 ) .
since the closure of the parameter space for @xmath54 does not depend on @xmath57 , the closure of the parameter space for bethe subalgebras in @xmath1 is @xmath3 as well .
next , we deduce from the results of shuvalov and tarasov that any limit of the algebras @xmath54 is a free polynomial algebra and a maximal commutative subalgebra in @xmath27 . since @xmath53 is asymptotically isomorphic to @xmath54 , the same is true for limit bethe subalgebras corresponding to points of @xmath3 ( theorem [ result ] ) . using shuvalov s description of limit shift - of - argument subalgebras
we explicitly describe the simplest limits corresponding to generic points of codimension @xmath14 strata in @xmath3 in terms of bethe subalgebras for smaller yangians and shift - of - argument subalgebras for smaller lie algebras ( theorem [ result2 ] ) .
iterating this we obtain the explicit description of all limit bethe subalgebras . to any semisimple lie algebra @xmath58 ( and even more generally , to any kac - moody algebra @xmath58 )
one can assign the yangian @xmath59 , the quantum group generated by the rational r - matrix of @xmath58 . to any element @xmath2 of the cartan torus @xmath8 of @xmath58 one
can assign a bethe subalgebra generated by traces of the products of @xmath2 with the r - matrix in all integrable representations .
this is a commutative subalgebra which is expected to be maximal for regular @xmath2 .
the variety parametrizing all possible limits of such commutative subalgebras is a compactification of the set @xmath11 of regular elements of the torus ( well - defined as a pro - algebraic scheme ) .
our theorem [ result ] states that in the case @xmath60 it is @xmath3 .
the natural generalization of this to lie algebras of other types is the de concini procesi closure @xcite of the complement to the following arrangement of subvarieties in a toric variety .
consider the toric variety @xmath61 ( acted on by the maximal torus @xmath62 ) which corresponds to the fan determined by the root hyperplanes .
equivalently , @xmath61 is the closure of a generic @xmath8-orbit in the flag variety @xmath63
. we can regard @xmath11 as a complement of an arrangement of hypersurfaces in @xmath61 .
following de concini and procesi @xcite , one can construct a compactification @xmath64 of @xmath11 by blowing up all indecomposable intersections of the hypersurfaces in @xmath61 .
@xmath64 is the parameter space for limit subalgebras of the family of bethe subalgebras in the yangian @xmath59 .
note that @xmath3 is the de concini procesi closure of @xmath11 for @xmath60 . for infinite root systems ( say , for affine kac - moody ) @xmath65 is not well - defined as an algebraic scheme but is still well - defined as a pro - algebraic scheme . on the other hand , in this case
the bethe subalgebras themselves live not in the yangian but rather in its completion which is an inverse limit of some quotients of the yangian .
so we can generalize our conjecture to yangians of infinite - dimensional lie algebras by stating that two inverse limits are isomorphic .
by analogy with the shift - of - argument algebras and gaudin algebras , we expect that for real values of the parameter the corresponding ( limit ) bethe algebra has simple spectrum in any irreducible finite - dimensional ( or integrable ) representation of the yangian .
following @xcite , one can assign a covering of @xmath66 to any irreducible representation of @xmath59 .
the fiber of such covering is just the set of joint eigenlines for the elements of a bethe algebra in this representation . for kirillov - reshetikhin modules ,
we expect a natural bijection of the fiber of this covering with the corresponding kirillov - reshetikhin crystal hence obtaining an action of the fundamental group @xmath67 on the crystal .
we also expect that this action can be described in purely combinatorial terms . in section 2 we recall some known facts about yangians and bethe subalgebras .
in section 3 we recall definition of moduli space @xmath68 . in section 4
we discuss some facts about shift of argument subalgebras .
section 5 contains main results . in section 6
we prove main theorems .
we thank michael finkelberg and alexander molev for helpful remarks and references .
this work has been funded by the russian academic excellence project 5 - 100. the work was finished during a.i.s internship at mit supported by nru hse .
is deeply indebted to mit and especially to r. bezrukavnikov and p. etingof , for providing warm hospitality and excellent working conditions .
the work of l.r . has been supported by the russian science foundation under grant 16 - 11 - 10160 .
we follow the notations and conventions of @xcite . _
yangian for @xmath4 _ is a complex unital associative algebra with countably many generators @xmath69 where @xmath70 , and the defining relations @xmath71 - [ t_{ij}^{(r)},t_{kl}^{(s+1 ) } ] = t_{kj}^{(r)}t_{il}^{(s ) } - t_{kj}^{(s)}t_{il}^{(r)},\ ] ] where @xmath72 and @xmath73 .
this algebra is denoted by @xmath1 .
it is convenient to consider the formal series @xmath74.\ ] ] we denote by @xmath75 the matrix whose @xmath76-entry is @xmath77 .
we regard this matrix as the following element of @xmath78 \otimes \rm{end } \
\mathbb{c}^n$ ] : @xmath79 where @xmath80 stands for the standard matrix units .
consider the algebra @xmath81 \otimes ( \rm{end } \
\mathbb{c}^n)^{\otimes n}.\ ] ] for any @xmath82 there is an embedding @xmath81 \otimes \rm{end } \
\mathbb{c}^n\to y({{\mathfrak{gl}}}_n)[[u^{-1 } ] ] \otimes ( \rm{end } \ \mathbb{c}^n)^{\otimes n}\ ] ] which is an identity on @xmath78 $ ] and embeds @xmath83 as the @xmath84-th tensor factor in @xmath85 .
denote by @xmath86 the image of @xmath75 under this embedding .
the symmetric group @xmath87 acts on @xmath78 \otimes ( \rm{end } \
\mathbb{c}^n)^{\otimes n}$ ] by permuting the tensor factors .
this action factors through the embedding @xmath88 hence the group algebra @xmath89 $ ] is a subalgebra of @xmath78 \otimes ( \rm{end } \ \mathbb{c}^n)^{\otimes n}$ ] .
let @xmath90 be the subgroup of @xmath87 permuting the first @xmath91 tensor factors .
denote by @xmath92 the antisymmetrizer @xmath93\subset y({{\mathfrak{gl}}}_n)[[u^{-1 } ] ] \otimes ( \rm{end } \
\mathbb{c}^n)^{\otimes n}.\ ] ] let @xmath2 be a @xmath94-valued @xmath95 matrix . for any @xmath82 denote by @xmath96 the element @xmath97 \otimes ( \rm{end } \
\mathbb{c}^n)^{\otimes n}$ ] . for
any @xmath98 introduce the series with coefficients in @xmath1 by @xmath99 where we take the trace over all copies of @xmath100 .
@xmath101 hence does not depend on @xmath2 up to proportionality .
the series @xmath102 is called quantum determinant and usually denoted by @xmath103 .
the following result is classical ( see e.g. @xcite ) : the coefficients of @xmath104 , @xmath105 pairwise commute .
the coefficients of @xmath103 generate the center of @xmath1 .
the commutative subalgebra generated by the coefficients of @xmath104 is called _
bethe subalgebra_. we denote it by @xmath0 . in @xcite nazarov and
olshanski proved the following theorem : [ maxfreetheorem ] suppose that @xmath106 has pairwise distinct nonzero eigenvalues .
then the subalgebra @xmath0 in @xmath1 is maximal commutative .
the coefficients of @xmath107 are free generators of @xmath0 . from now on
we suppose @xmath2 is diagonal with the eigenvalues @xmath108 .
let us write @xmath104 explicitly in the generators @xmath77 : [ vid ] we have @xmath109 where @xmath110 are the quantum minors .
note that @xmath111 , where @xmath112 it is clear that in the basis @xmath113 with the standard order we have @xmath114 then in the same basis we have @xmath115 note that if there exists @xmath116 then @xmath117 .
suppose that @xmath118 for @xmath119 .
then @xmath120 this implies that the trace is equal to @xmath121 the last equation follows from skew - symmetricity of quantum minors .
we define an ascending filtration on @xmath1 by setting the degree of the generators as @xmath122 denote by @xmath123 the vector subspace consisting of all elements with degrees not greater than @xmath124 .
consider the associated graded algebra [ filtration ] @xmath125 there is the following analog of pbw - theorem for @xmath1 ( see @xcite ) : [ pbw]@xmath126 is the algebra of polynomials in the variables @xmath127 , where @xmath127 is the image of @xmath128 in @xmath129 .
the polynomial algebra @xmath126 has a natural poisson structure .
for any two elements @xmath130 of degrees @xmath131 respectively the poisson bracket of their images @xmath132 in @xmath126 is @xmath133 \mod y_{p+q-2}({{\mathfrak{gl}}}_n).\ ] ] let us define two different embedding of @xmath1 to @xmath134 : @xmath135 @xmath136 according to pbw - theorem these maps are injective .
define a homomorphism @xmath137 there @xmath138 is a standard matrix unit .
@xmath139 is a surjective homomorphism from @xmath1 to @xmath16 known as _ evaluation _ homomorphism . by definition ,
put @xmath140 it is well - known that @xmath141 is an involutive automorphism of @xmath1 .
we define a homomorphism @xmath142 note that @xmath143 is injective .
[ ipsi ] ( * ? ? ?
* proposition 1.13.1 ) @xmath144 is centralized by @xmath145 in @xmath134 .
we will need a bit more precise statement : [ commuteyang ] the homomorphisms @xmath146 and @xmath147 define an embedding @xmath148 .
it suffices to show that the associated graded @xmath149 is injective .
but the latter is sends @xmath150 to @xmath128 and @xmath151 to @xmath152 , where @xmath153 is an expression of of @xmath154 with @xmath155 .
hence it is clearly injective .
consider the map @xmath156 it is clear that @xmath157 .
let @xmath158 $ ] be the polynomial algebra of infinite many variables .
define a grading on @xmath159 by setting @xmath160 . for any @xmath57 we have a surjective homomorphism @xmath161 where @xmath162 , @xmath163 are some generators of @xmath164 of degree @xmath165 , see @xcite .
consider the algebra @xmath166 .
this algebra has a well - defined ascending filtration given by @xmath167 for any @xmath168 we define homomorphisms of filtered algebras @xmath169 denote by @xmath170 the @xmath26-th filtered component , i.e. the vector space of the elements of degree not greater than @xmath26 . from (
* theorem 8.4.3 ) we have : [ qasym ] the sequence @xmath171 is an asymptotical isomorphism .
this means that for any @xmath26 there exists @xmath172 such that for any @xmath51 the restriction of @xmath173 to the @xmath26-th filtered component @xmath170 is an isomorphism of vector spaces @xmath174 . according to lemma [ commuteyang ]
, there is a tensor product of two commuting yangians @xmath175 and @xmath176 inside @xmath1 .
let us see what happens with these two commuting yangians under the centralizer construction map @xmath177 .
the algebra @xmath56 is generated by @xmath178 ; @xmath179 is generated by @xmath180 .
the subalgebra @xmath181 is generated by @xmath182 .
hence @xmath183 and @xmath184 are naturally subalgebras in @xmath27 .
the restriction of @xmath173 to @xmath185 is @xmath186 .
the restriction of @xmath173 to @xmath187 is @xmath188 .
the first statement is immediate from the following commutative diagram : @xmath189{e , t}{i_{m } } \arrow{ese , b}{i_{k+m } } \node[2]{y({{\mathfrak{gl}}}_n ) } \arrow{s , r}{i_k}\\ \node[3]{y({{\mathfrak{gl}}}_{n+k } ) } \end{diagram}\ ] ] to prove second statement , consider another diagram : @xmath190 the lower square is commutative by definition so it is enough to prove that upper square is commutative as well .
but it follows from ( * ? ? ?
* lemma 1.11.2 ) .
let @xmath191 be a cartan subalgebra of diagonal matrices , @xmath11 regular elements of @xmath192 .
it is possible to construct new commutative subalgebras as limits of bethe subalgebras .
we describe what the `` limit '' means here .
let @xmath2 be an element of @xmath11 .
consider @xmath193 .
in @xcite it was proved that the images of the coefficients of @xmath107 freely generate the subalgebra @xmath194 .
hence the dimension @xmath195 of @xmath196 does not depend on @xmath2 .
therefore for any @xmath197 we have a map @xmath198 from @xmath11 to @xmath199 such that @xmath200 .
denote the closure of @xmath201 ( with respect to zariski topology ) by @xmath202 .
there are well - defined projections @xmath203 for all @xmath204 .
let @xmath205 be a sequence where @xmath206 such that @xmath207 . from this sequence
we obtain a sequence @xmath208 where @xmath209 is a subspace of @xmath210 such that @xmath211 .
let us define subalgebra @xmath212 .
we claim that this subalgebra is commutative and we call it _ limit subalgebra_. indeed , @xmath213 is a commutative subalgebra because being a commutative subalgebra is a zariski - closed condition .
note that this subalgebra has the same poincare series as @xmath0 .
further let us define inverse limit @xmath214 .
this limit is well - defined as a pro - algebraic scheme and parametrizes all limit bethe subalgebras . in @xcite
the limits shift of argument subalgebras are defined in the analytic topology , just as limits of @xmath14-parametric families of generic shift of argument subalgebras .
but it is well - known that the closure of an affine algebraic variety in a complex projective space with respect to zariski topology coincides with its closure with respect to the analytic topology .
so the closure of the parameter space is the same for both definitions of the limit . in our work
we use both approaches .
let @xmath68 denote the deligne - mumford space of stable rational curves with @xmath215 marked points .
the points of @xmath68 are isomorphism classes of curves of genus @xmath216 , with @xmath215 ordered marked points and possibly with nodes , such that each component has at least @xmath217 distinguished points ( either marked points or nodes ) .
one can represent the combinatorial type of such a curve as a tree with @xmath215 leaves with inner vertices representing irreducible components of the corresponding curve , inner edges corresponding to the nodes and the leaves corresponding to the marked points .
informally , the topology of @xmath68 is determined by the following rule : when some @xmath57 of the distinguished points ( marked or nodes ) of the same component collide , they form a new component with @xmath218 distinguished points ( the new one is the inersection point with the old component ) . in particular , the tree describing the combinatorial type of the less degenerate curve is obtained from the tree corresponding to the more degenerate one by contracting an edge .
the space @xmath68 is a smooth algebraic variety .
it can be regarded as a compactification of the configuration space @xmath219 of ordered @xmath220-tuples @xmath221 of pairwise distinct points on @xmath222 modulo the automorphism group @xmath223 .
since the group @xmath223 acts transitively on triples of distinct points , we can fix the @xmath220-th point to be @xmath224 and fix the sum of coordinates of other points to be zero .
then the space @xmath219 gets identified with the quotient @xmath225 where @xmath226 , and the group @xmath227 acts by dilations . under this identification of @xmath219 ,
the space @xmath68 is just the git quotient of the iterated blow - up of the subspaces of the form @xmath228 in @xmath229 by the natural @xmath227 action by dilations .
the space @xmath68 is stratified as follows .
the strata are indexed by the combinatorial types of stable rational curves , i.e. by rooted trees with @xmath26 leaves colored by the marked points @xmath230 ( the root is colored by @xmath231 ) .
the stratum corresponding to a tree @xmath8 lies in the closure of the one corresponding to a tree @xmath232 if and only if @xmath232 is obtained from @xmath8 by contracting some edges .
the spaces @xmath233 form a topological operad .
this means that one can regard each point of the space @xmath233 as an @xmath26-ary operation with the inputs at marked points @xmath230 and the output at @xmath234
. then one can substitute any operation of this form to each of the inputs .
more precisely , for any partition of the set @xmath235 into the disjoint union of subsets @xmath236 with @xmath237 there is a natural substitution map @xmath238 which attaches the @xmath165-th curve @xmath239 to the @xmath165-th marked point of the curve @xmath240 by gluing the @xmath241-th marked point of each @xmath242 with the @xmath165-th marked point of @xmath243 .
in fact all substitution maps @xmath244 are compositions of the elementary ones with @xmath245 .
the compositions of the substitution maps are indexed by rooted trees describing the combinatorial type of the ( generic ) resulting curves .
in particular , each stratum of @xmath233 is just the image of the product of the open strata of appropriate @xmath246 under some composition of substitution maps . in particular
, strata of codimension @xmath14 are just the images of the open strata of @xmath247 under the elementary substitution maps and can be obtained as the limit set when the points @xmath248 collide and other points stay isolated . since
each substitution map is a composition of the elementary ones , generic point of each stratum can be obtained iterating this limiting procedure .
let @xmath58 be a reductive lie algebra . to any
@xmath249 one can assign a poisson - commutative subalgebra in @xmath250 with respect to the standard poisson bracket ( coming from the universal enveloping algebra @xmath251 by the pbw theorem ) .
let @xmath252 be the center of @xmath250 with respect to the poisson bracket .
the algebra @xmath253 generated by the elements @xmath254 , where @xmath255 , ( or , equivalently , generated by central elements of @xmath256 $ ] shifted by @xmath34 for all @xmath35 ) is poisson - commutative and has maximal possible transcendence degree .
more precisely , we have the following [ mf ] @xcite for regular semisimple @xmath257 the algebra @xmath36 is a free commutative subalgebra in @xmath250 with @xmath258 generators ( this means that @xmath36 is a commutative subalgebra of maximal possible transcendence degree ) .
one can take the elements @xmath259 , @xmath260 , @xmath261 , where @xmath262 are basic @xmath58-invariants in @xmath250 , as free generators of @xmath36 .
in particular , for @xmath263 we have : + @xmath26 generators of degree 1 ; + @xmath264 of degree @xmath265 ; + + 1 of degree @xmath26 .
let @xmath266 be a cartan subalgebra of the lie algebra @xmath58 .
denote by @xmath267 the set of positive roots of @xmath58 .
we assume that @xmath2 is a regular semisimple element of the cartan subalgebra @xmath268 .
the linear and quadratic part of the subalgebras @xmath36 can be described as follows ( see @xcite ) : @xmath269 the results of vinberg @xcite and shuvalov @xcite imply that the limit subalgebra of the family @xmath36 is uniquely determined by its quadratic component .
hence the variety parametrizing limit shift of argument subalgebras in @xmath250 is just the closure of the family @xmath39 in the grassmannian @xmath270 . in @xcite shuvalov
described the closure of the family of subalgebras @xmath253 under the condition @xmath15 ( i.e. , for regular @xmath2 in the fixed cartan subalgebra ) .
in particular , the following holds : [ shuvalov]@xcite suppose that @xmath271 for generic @xmath272 .
set @xmath273 ( where @xmath274 is the centralizer of @xmath242 in @xmath58 ) , @xmath275 .
then we have 1 . the subalgebra @xmath276 is generated by all elements of @xmath277 and their derivatives ( of any order ) along @xmath278 for all @xmath57 .
2 . @xmath279 is a free commutative algebra .
one can take the generators of @xmath280 for all @xmath57 and some of their derivatives along @xmath278 as a set of free generators of the limit subalgebra @xmath279 .
moreover , according to the results of a. tarasov @xcite the subalgebras dicussed above ( both @xmath36 and the limit ones ) are maximal poisson - commutative subalgebras in @xmath250 ( i.e. coincide with their poisson centralizers ) .
in @xcite the subalgebras @xmath36 were quantized , i.e. the existence of such commutative subalgebras @xmath281 that @xmath282 was proved . in @xcite
it is proved that @xmath36 is the poisson centralizer of the space @xmath39 for generic @xmath2 hence the lifting of @xmath36 to @xmath251 is unique for generic @xmath2 . in the case
@xmath263 we have a stronger statement due to a. tarasov @xcite : any subalgebra from this family ( including the limit ones ) can be uniquely lifted to the universal enveloping algebra and , moreover , there is a particular choice of the generators such that this lifting is just the symmetrization map on the generators . in particular , for @xmath263 , the varieties parametrizing limit subalgebras of the families @xmath36 and @xmath31 are the same . in fact , we expect that the latter is true for arbitrary @xmath58 . in the case
@xmath263 the subalgebra @xmath283 is in fact the image of @xmath284 under the evaluation homomorphism @xmath285 ( we can assume that @xmath2 is nondegenerate since @xmath31 does not change after adding a scalar matrix to @xmath2 ) .
this fact plays crucial role in what follows .
unfortunately this does not generalize to arbitrary @xmath58 since there is no evaluation homomorphism in general .
in @xcite the explicit description of limit subalgebras is given together with a set - theoretical description of the parameter space .
the results of shuvalov in the @xmath286 case can be reformulated in the following way .
the shift of argument subalgebras in @xmath287 depend on a parameter @xmath15 and do not change under the transformations @xmath288 . so the parameter space for the subalgebras @xmath289 can be regarded as @xmath219 .
[ prop - fclosure ] the closure of the parameter space for the shift - of - argument subalgebras in @xmath287 is @xmath233 .
the same is true for the parameter space for the family @xmath31 .
for @xmath263 the regular cartan element @xmath2 is a diagonal matrix with pairwise distinct eigenvalues @xmath230 . from the results of shuvalov @xcite and vinberg
@xcite it follows that any limit subalgebra is uniquely determined by its quadatic graded component . according to the results of tarasov @xcite ,
the lifting of any limit shift - of - argument subalgebra is unique hence uniquely determined by the quadratic component as well .
note that in this case @xmath39 is the linear span of the quadratic elements @xmath37 which are the coefficients of ( an appropriate version of ) the @xmath38 connection .
in particular the space @xmath39 does not change under simultaneous affine transformations of the @xmath41 s , hence the space of parameters is naturally the configuration space of @xmath18 pairwise distinct points on the affine line or , equivalently , the configuration space @xmath42 of @xmath43 pairwise distinct points on the projective line . in @xcite aguirre ,
felder and veselov considered the same subspaces @xmath39 universally ( i.e. as subspaces in the drinfeld - kohno holonomy lie algebra ) and showed that the closure of the family @xmath39 is the deligne - mumford compactification @xmath68 .
this means that the parameter space for limit shift of argument subalgebras ( both classical and quantum ) for @xmath263 is @xmath68 . one can define the subalgebra @xmath290 corresponding to a degenerate curve @xmath291 recursively as follows .
let @xmath292 be the irreducible component of @xmath61 containing the marked point @xmath234 . to any distinguished point @xmath293
we assign the number @xmath294 of marked points on the ( reducible ) curve @xmath295 attached to @xmath292 at @xmath296 .
let @xmath2 be the diagonal matrix with the eigenvalues @xmath296 of multiplicity @xmath294 for all distinguished points @xmath293 .
then the corresponding shift of argument subalgebra @xmath289 is centralized by the lie subalgebra @xmath297 in @xmath286 and contains the poisson center @xmath298 .
the subalgebra corresponding to the curve @xmath61 is just the product of @xmath289 and the subalgebras corresponding to @xmath295 in @xmath299 for all distinguished points @xmath293 ( for this we need to define the point @xmath234 on each @xmath295 it is just the intersection with @xmath292 ) .
having a unique lifting of each @xmath36 to the universal enveloping algebra @xmath25 we can do the same and attach a commutative subalgebra in @xmath25 to any stable rational curve @xmath300 .
since every limit shift of argument subalgebra has a unique lifting to @xmath16 , every limit subalgebra of the family @xmath283 is of this form .
now we can make a precise statement : [ tensorprod ] the subalgebra @xmath301 corresponding to a degenerate curve @xmath61 is the tensor product @xmath302 .
the same is true for the lifting @xmath303 . in order to prove this proposition we need the following lemma : [ tensor ] suppose that @xmath58 is a reductive lie algebra , @xmath304 reductive subalgebra of @xmath58 .
then the subalgebras @xmath305 and @xmath306 in @xmath251 are both free @xmath307-modules .
moreover , the product of these subalgebras in @xmath251 is : @xmath308 the same holds for the associated graded algebras .
namely , @xmath309 and @xmath310 are free @xmath311-modules , and @xmath312 we can assume without loss of generality that @xmath304 does not contain nontrivial ideals of @xmath58 .
then this is a particular case of knop s theorem on harish - chandra map for reductive group actions ( see @xcite , theorem 10.1 and items ( d ) and ( e ) of the main theorem ) . indeed , the main theorem of @xcite states that for any reductive @xmath313 and any smooth affine @xmath313-variety @xmath61 the algebra @xmath314 of @xmath313-invariant differential operators and its commutant @xmath315 in the algebra @xmath316 are both free modules over the center of @xmath314 .
moreover , the product @xmath317 is the tensor product @xmath318 of @xmath314 and @xmath315 over the center of @xmath314 . to get the desired statement
we just apply this to the @xmath319-action on @xmath320 , where @xmath321 acts from the right and @xmath322 acts on from the left . for this case
we have @xmath323 is @xmath305 and its commutant in @xmath316 is @xmath324 ( generated by momenta of the left @xmath322-action and the right @xmath321-action ) . by theorem 10.1 of @xcite , the center of @xmath305 is @xmath325 and the algebra @xmath315 contains @xmath326 as the subalgebra of ( right ) @xmath321-invariants .
hence by the main theorem of @xcite we have the desired assertion .
theorems 9.4 and 9.8 of @xcite imply the same for the associated graded algebras ( in fact the main theorem of @xcite is a consequence of the same fact for the associated graded algebras ) .
now let us describe the simplest limit subalgebras corresponding to the case when all the curves @xmath295 are irreducible .
[ size ] let @xmath327 and @xmath328 for @xmath329 such that @xmath330 and @xmath331 for @xmath332 .
then the element @xmath333 is regular for small @xmath272 .
the subalgebra @xmath334 is the tensor product @xmath335 here @xmath336 is a subalgebra in @xmath337 .
moreover , @xmath338 .
the same holds for quantum shift of argument subalgebras @xmath31 . from @xcite
it follows that @xmath339 .
we know that @xmath340 ( because @xmath341 = \partial_{[c , x]}$ ] ) and @xmath342 .
moreover , from proof of ( * ? ? ?
* lemma 1 ) it follows that @xmath343 . using lemma [ tensor ]
we obtain the result .
since the closure of the parameter space for subalgebras @xmath36 is @xmath45 , any limit subalgebra can be obtained by iterating the above degeneration .
iterating limits from lemma above we obtain proposition [ tensorprod ] .
lemma [ size ] implies the following formula for the poincare series of @xmath36 for any diagonal @xmath2 .
denote by @xmath344 the formal power series @xmath345 .
we set @xmath346 .
note that @xmath347 is the poincare series of @xmath36 for regular @xmath2 .
let @xmath348 such that @xmath330 .
then @xmath36 is a free polynomial algebra with the poincare series @xmath349 straightforward from lemma [ size ] .
[ l36 ] let @xmath350 be a family of subalgebras of the form @xmath351.then 1 ) every limit subalgebra of the family @xmath350 is a maximal commutative subalgebra of @xmath352 .
+ 2 ) every limit subalgebra of this family is a free polynomial algebra with the following poincare series : @xmath353 3 ) the closure of the parameter space for the family @xmath350 is @xmath354 .
let @xmath355 belongs to @xmath11 for small @xmath272 and consider @xmath356 .
let @xmath357 and degree @xmath26 such that @xmath358 and @xmath359 is regular for small @xmath272 .
then @xmath360 is free and maximal commutative .
suppose that @xmath361 is not maximal .
this implies that there exists an element @xmath362 such that @xmath363 .
this means that @xmath364 commutes with every element from the limit subalgebra ( [ f1 ] ) .
but the limit subalgebra is maximal therefore @xmath365 which is @xmath361 by lemma [ size ] .
thus @xmath361 is maximal .
the second assertion follows from the fact that poincare series of limit subalgebra + @xmath361 is the same as for algebra @xmath366 for generic @xmath272 and from lemma [ fsize ] . to prove third assertion let fix @xmath367 and consider a family of subalgebras of the form @xmath368 where @xmath369 .
this family is parameterized by + @xmath370 .
but the closure of @xmath371 in @xmath372 is @xmath373 therefore the closure of the parameter space for the family @xmath350 is @xmath354 .
the same results ( lemmas [ size ] , [ fsize ] and [ l36 ] ) are true for the liftings of shift of argument subalgebras to the universal enveloping algebra @xmath25
. follows from the fact that shift of argument subalgebras admits unique lifting to @xmath25 + @xcite .
by analogy with the shift of argument subalgebras , we can regard the space of parameters for the family of bethe subalgebras in the yangian , @xmath19 as the moduli space @xmath20 , i.e. the space of rational curves with @xmath21 marked points .
more precisely , to any matrix @xmath2 with the eigenvalues @xmath22 , we can assign a rational curve @xmath23 with the marked points @xmath24 .
let @xmath374 be the closure of this parameter space in the sense of [ def - lim ] .
our main result is the following [ result ] 1 .
@xmath374 is a smooth algebraic variety isomorphic to @xmath3 .
2 . for any point @xmath375 , the corresponding commutative subalgebra @xmath376 in @xmath1 is free and maximal commutative .
next , the operad structure on the spaces @xmath233 leads to a recursive description of the limit subalgebras analogous to that of the limit shift - of - argument subalgebras .
let @xmath292 be the irreducible component of @xmath377 containing the marked point @xmath234 .
we identify @xmath292 with the standard @xmath378 in such a way that the marked point @xmath234 is @xmath234 and the point where the curve containing the marked point @xmath216 touches @xmath292 is @xmath216 . to any distinguished point @xmath293
we assign the number @xmath294 of marked points on the maximal ( possibly reducible ) curve @xmath295 attached to @xmath292 at @xmath296 ( we set @xmath379 if @xmath295 is a ( automatically , marked ) point ) .
let @xmath2 be the diagonal @xmath380-matrix with the eigenvalues @xmath296 of multiplicity @xmath294 for all distinguished points @xmath381 .
then the subalgebra @xmath382 centralized by the lie subalgebra @xmath383 in @xmath384 and by the complement sub - yangian @xmath385 .
the subalgebras corresponding to boundary points of @xmath354 can be inductively described as follows : [ result1 ] 1 . the limit bethe subalgebra corresponding to the curve @xmath377 is the product of the following @xmath217 commuting subalgebras : first , @xmath386 , second , the subalgebra corresponding to @xmath387 in the complement sub - yangian @xmath388 and third , the limit shift - of - argument subalgebras @xmath389 in @xmath390 for all distinguished points @xmath391 ( again we define the point @xmath234 on each @xmath295 just as the intersection with @xmath292 )
@xmath382 contains the center of every @xmath392 .
the above product is in fact the tensor product @xmath393 [ result2 ] 1 ) let @xmath394 and @xmath395 .
suppose that @xmath396 ( i.e. both @xmath243 and @xmath397 are regular and non - degenerate ) .
then @xmath398 + 2 ) let @xmath327 and @xmath328 for @xmath329 such that @xmath330 and @xmath331 for @xmath332 .
let @xmath399 then @xmath400 where @xmath401 is the shift of argument subalgebra in @xmath402 ( the copy of @xmath16 in the yangian @xmath1 is generated by @xmath403 ) .
theorems [ result1 ] and [ result2 ] are equivalent .
indeed , theorem [ result2 ] is a particular case of theorem [ result1 ] . on the other hand
, any degenerate curve @xmath61 can be obtained from a non - degenerate one by , first , taking a limit from part 1 of theorem [ result2 ] ( @xmath91 times where @xmath91 is the number of irreducible components of @xmath61 on the way between the marked points @xmath216 and @xmath234 ) and , second , taking a limit from part 2 of theorem [ result2 ] for each of that @xmath91 components .
this reduces the situation to shift of argument subalgebras .
so theorem [ result1 ] follows from theorem [ result2 ] .
let @xmath404 be a non - regular element from the maximal torus .
let @xmath405 be the number of homogeneous degree @xmath165 generators of @xmath36 ( i.e. the multiplicity of the factor @xmath406 in the poincare series of @xmath36 ) .
it is enough to check that in @xmath409 we have enough algebraically independent generators ( see @xmath410 ) .
the images of @xmath411 are @xmath412 now we can change generators @xmath413 .
note that there is a filtration in @xmath126 given by @xmath414 according to this filtration the leading term of the coefficient of @xmath415 at @xmath416 for @xmath417 is @xmath418 here we set @xmath419 we see that , for @xmath420 the leading terms of such coefficients do not depend on @xmath2 if we change generators to @xmath421 .
moreover , we see that the leading terms of the coefficients of @xmath422 at @xmath423 consists of only @xmath424 , coefficients of @xmath425 of @xmath415 contain @xmath426 and do not contain @xmath427 and so on . generally ,
we have the following to see that the set of coefficients of @xmath428 of @xmath415 are algebraically independent let us compute differential at point @xmath437 , @xmath438 .
it is straightforward computation that @xmath439 hence the differential of the coefficient of @xmath415 at @xmath428 is a non - zero linear combination of @xmath440 with @xmath441 .
hence differentials of all coefficients with @xmath417 are linearly independent .
therefore if we have algebraic dependence between some coefficients , then it can only consist of coefficients @xmath416 with @xmath442 .
m1.2cm| & @xmath443 & @xmath444 & @xmath445 & & @xmath446 & @xmath447 & @xmath448 & + @xmath449 & 1 & 2 & 3 & & @xmath450 & @xmath18 & @xmath43 & + @xmath451 & 1 & 1 & 2 & & @xmath452 & @xmath450 & @xmath18 & + @xmath453 & 1 & 1 & 1 & & @xmath454 & @xmath452 & @xmath450 &
+ @xmath455 & & & & & & & & + @xmath456 & 1 & 1 & 1 & & 1 & 1 & 2 &
+ on the other hand , the coefficients of @xmath431 of @xmath411 , @xmath433 generate the shift of argument subalgebra @xmath36 in @xmath33 hence we have the desired number of algebraically independent generators .
note that the series @xmath461 is invertible so multiplying by @xmath461 is a bijection on a set of generators of @xmath0 .
moreover @xmath462 is invertible automorphism of @xmath1 so we can consider @xmath463 instead of @xmath104 as a generators of @xmath0 . by lemma [ togd ] @xmath464 on the other hand , from lemma [ betheimage ] we have the following : @xmath468 so the limit subalgebra contains @xmath469 .
the latter is @xmath470 due to lemma [ commuteyang ] .
since @xmath471 and @xmath397 are regular elements of @xmath472 and @xmath473 respectively , the tensor product has @xmath470 has the same poincare series as generic bethe algebra , therefore the limit is in fact equal to @xmath470 .
so theorem [ result2 ] ( 1 ) is proved .
[ incl ] let @xmath10 be any non - degenerate diagonal matrix .
set @xmath474 .
then @xmath475 for any non - degenerate @xmath476 and @xmath477 .
moreover , the restrictions of @xmath171 to @xmath0 give an asymptotic isomorphism of @xmath0 and @xmath478 , @xmath479 . the same is true for all limit algebras of the family @xmath0 .
we have @xmath480 .
consider @xmath481 and subalgebra @xmath482 in @xmath134 .
then by theorem [ result2 ] ( 1 ) the limit subalgebra @xmath483 is equal to @xmath484 .
note that @xmath485 .
we have @xmath486 we use here the fact that @xmath487 , see @xcite . but
@xmath488 therefore @xmath489 .
the fact that @xmath171 is asymptotic isomorphism follows from lemma [ bethesize ] . for limit algebras it follows from the fact that @xmath490 and the fact that limit algebras have the same poincare series .
analogously to [ pf1 ] , we have @xmath491 in limit subalgebra . also we know that @xmath492 we see that @xmath493 maps to @xmath494 .
moreover,@xmath173 is asymptotic isomorphism and restriction of @xmath173 to @xmath495 is identity , so we have @xmath496 in the limit subalgebra as well . but
using proposition [ bethesize ] we see that the poincare series of @xmath497 coincides with that of @xmath498 for generic @xmath272 . from proposition [ incl ]
we know that family \{@xmath173 } is asymptotic isomorphism between @xmath499 and @xmath500 . according to lemma [ l36 ] all algebras @xmath500 form the family @xmath350 .
limit subalgebras of this family are parameterized by @xmath354 for any @xmath501 .
therefore limit bethe subalgebras are parameterized by @xmath354 as well .
additionally , all limit algebras of the family @xmath350 are free and maximal commutative by lemma [ l36 ] .
therefore all limit bethe subalgebras in the yangian are free and maximal commutative as well . we have already proved the first part of the theorem .
we only need to see that @xmath382 contains the center of every @xmath392 .
we know that the image @xmath502 contains the center of every @xmath503 for any @xmath57 .
also we know that @xmath159 maps to the center of @xmath504 by @xmath173 .
but from ( * ? ? ?
* lemma 10.1 ) it follows that the generators of the center of @xmath504 and of the centers of all @xmath503 are algebraically independent . using last time the fact that @xmath173 is asymptotic isomorphism we obtain the result . | we study degenerations of bethe subalgebras @xmath0 in the yangian @xmath1 , where @xmath2 is a regular diagonal matrix .
we show that closure of the parameter space of the family of bethe subalgebras is the deligne - mumford moduli space of stable rational curves @xmath3 and state a conjecture generalizing this result to bethe subalgebras in yangians of arbitrary simple lie algebra .
we prove that all subalgebras corresponding to the points of @xmath3 are free and maximal commutative .
we also describe explicitly the `` simplest '' degenerations and show that every degeneration is the composition of the simplest ones . |
in recent years , our understanding of pulsations of non - rotating relativistic stars has been much improved . in particular , the role of gravitational wave in the stars becomes much clear .
there exists the oscillation mode named as w - mode ( gravitational _ w_ave mode ) , associated with the gravitational wave .
the gravitational wave is inherent in the general relativity , so that the mode becomes evident only for relativistic system . except for this new mode ,
the relativity little affects the modes known in the newtonian pulsation theory .
the general relativity slightly changes the oscillation frequency and gives rise to a very slow damping of the mode .
the emission of gravitational radiation implies that the oscillation frequency should be complex with a relatively tiny imaginary part .
( see andersson , kojima , and kokkotas and references therein for the present status of the oscillation spectra of non - rotating relativistic stars . )
the perturbation of the gravity is not so important , if the w - mode and the decay of the pulsations due to the gravitational radiation do not matter .
this fact hints at the further simplification of the pulsation problem , that is , neglecting the perturbation of the gravitational field .
this approximation is known as the cowling approximation ( cowling ) in the newtonian stellar pulsation theory , and gives the same qualitative results and reasonable accuracy of the oscillation frequencies .
( see , e.g. , cox . ) two different prescriptions , so far , had been proposed for the relativistic cowling approximation in non - radial pulsations of the non - rotating stars .
one method is that all metric perturbations are neglected , i.e. , @xmath1 @xcite .
the other is that the @xmath2 component of the metric perturbations is retained in the pulsation equations @xcite .
lindblom and splinter examined the accuracy of these two versions of the approximation for the dipole p - modes , and concluded that the mcdermott , van horn , and scholl version is more accurate than finn s version .
however , the finn s version is superior in g - mode calculations @xcite .
in contrast to the oscillations in the non - rotating stars , the calculation of oscillation frequencies in rapidly rotating relativistic stars seems to be very difficult task .
the equations become significantly complicated due to the relativity and rotation .
see priou for a set of the explicitly written , lengthy equations .
therefore , the calculation of normal frequencies was limited to the non - rotating and slowly rotating stars at best @xcite .
the possibility to use the cowling approximation in rotating relativistic system should be studied .
ipser and lindblom formulated the cowling approximation of the relativistic system , setting the metric perturbation to zero .
if the approximation is applied to the stellar pulsation , the equation becomes single second order partial differential equation .
the equation is manageable and will be hopefully solved in the future .
the crucial point is how good the cowling approximation is to estimate the oscillation frequency in the relativistic regime . in this paper
, we will estimate the accuracy of the cowling approximation in the frequencies of the non - rotating star and its rotational corrections .
that is , we will compare the frequencies of the cowling approximation with those of the relativistic perturbation theory . in section 2 , we present the formalism to calculate normal frequencies and the rotational shifts by the the cowling approximation . in section 3 , the comparison between both numerical results is given .
finally , section 4 is devoted to the discussion . throughout this paper
we will use units of @xmath3 .
to assess the accuracy of the cowling approximation , two distinct calculations are performed . in one calculation , the relativistic perturbation equations are solved to find normal frequencies and the rotational corrections for the pulsation of stellar models .
the equations and techniques are described in kojima ( 1992,1993 ) . in the other calculation
, we adopt the relativistic cowling approximation .
we restrict ourselves to the slowly rotating case , and write down the equations of the pulsation and rotational corrections below .
normal frequencies are real numbers , because gravitational perturbations are neglected in the cowling approximation . on the other hand ,
the normal frequencies are complex numbers with small imaginary part representing the damping of the gravitational radiation , when solving the exact pulsation equations .
we make a comparison in the oscillation frequency only .
we assume that the star is slowly rotating with a uniform angular velocity @xmath4 . in this paper
we consider only the first order effect with @xmath5 , where @xmath6 and @xmath7 are the radius and the mass of the star , respectively .
the geometry of a slowly rotating star in general relativity is described by @xmath8 where @xmath9 , @xmath10 and @xmath11 are functions of @xmath12 only .
given an equation of state @xmath13 the construction of the slowly rotating equilibrium model is well - known .
( for details see , e.g. , hartle ; chandrasekhar and miller . )
we will assume a polytropic equation of state , @xmath14 where @xmath15 and @xmath16 are constants .
the function @xmath17 is of order @xmath18 and the others , @xmath19 , @xmath20 , @xmath21 , and @xmath22 are the same as in the non - rotating case . in this subsection , we will explicitly write down the pulsation equation for the slowly rotating stars up to the first order of @xmath23 , using the equations derived by ipser and lindblom . we consider only the adiabatic perturbation for the perfect fluid , and use the simplest model for the adiabatic index @xmath24 based on the structure of the equilibrium configuration , @xmath25 the frequencies of the all @xmath26-modes are therefore zero in this case .
the perturbations are given in terms of @xmath27 , radial function @xmath28 , and spherical harmonics @xmath29 , where the azimuthal dependence of the spherical harmonics is @xmath30 in our convention .
the oscillation frequency @xmath31 is the value measured in the inertial frame .
corresponding to equations ( 16)(18 ) of ipser and lindblom , the eulerian changes in the pressure , the density , and the four - velocity are written up to the first order of @xmath23 as @xmath32 @xmath33 @xmath34 @xmath35 \ , y_{lm } \ , e^{-i \sigma t } \ , \label{def4}\end{aligned}\ ] ] @xmath36 \ , e^{-i \sigma t } \ ,
\label{def5}\ ] ] @xmath37 \ , e^{-i \sigma t } \ , \label{def6}\end{aligned}\ ] ] where @xmath38 the terms with @xmath39 are rotational corrections to the oscillation in the non - rotating star .
since the background geometry is spherically symmetric , the angular part can be separated into each @xmath40 mode .
the basic equation to be solved becomes second order ordinary differential equation for @xmath41 .
( see equation ( 38 ) of ipser and lindblom . )
we have @xmath42 \delta v = m \ , ( 2 \sigma \ , \omega \ , c - s ) \ , \delta v \ , \label{basiceq1}\ ] ] where @xmath43 @xmath44 - l(l+1 ) \ , ( \rho + p ) \ , e^{(\nu+\lambda)/2 } \ , \label{basiceq3}\ ] ] @xmath45 @xmath46 \biggr ] \
, \nonumber \\ \label{basiceq5}\end{aligned}\ ] ] where @xmath47 and @xmath48 are of the order @xmath23 .
we will solve the eigen - value problem , equation ( [ basiceq1 ] ) by standard perturbation method , that is , expand the potential @xmath41 and the frequency @xmath49 with respect to the small parameter @xmath23 as , @xmath50 @xmath51 the equation of order @xmath52 is for the non - rotating case , @xmath53 \delta v_0 = 0 \ , \label{basiceq11}\ ] ] and the equation of order @xmath54 is @xmath55 \delta v_1 + 2 \sigma_0 \sigma_1 \ , c \ , \delta v_0 = & & \nonumber \\
m \sqrt { { m \over \omega^2 r^3 } } \ , ( 2 \sigma_0 \ , \omega \ , c - s_0 ) \ , \delta v_0 \ , & & \label{basiceq12}\end{aligned}\ ] ] where @xmath56 is defined by equation ( [ basiceq5 ] ) with @xmath57 multiplying @xmath58 to equation ( [ basiceq12 ] ) , and integrating over the star , we have the first order rotational correction of the normal frequency @xmath59 as @xmath60 \ . \label{defsigma1}\ ] ] similar expression can be found in the newtonian pulsation equation , but is written by the lagrangian displacement , not by the eulerian change of the potential .
note also that the correction ( [ defsigma1 ] ) can be determined from the eigen - function and eigen - value in the non - rotating case , i.e. , @xmath61 and @xmath62 .
therefore , it is sufficient to calculate @xmath61 and @xmath62 . the frequency in the cowling approximation should be compared with the real part of eigen - frequency calculated by the exact pulsation equation .
the frequency @xmath49 of the slowly rotating star can be parameterized as @xmath63 where @xmath64 the subscript @xmath6 means the real part of the exact normal frequency .
we will compare results of two calculations in @xmath65 and @xmath66 .
as mentioned in the last subsection , we need eigen - frequency and eigen - function of the pulsations in the non - rotating stars .
the method to solve equation ( [ basiceq11 ] ) is well - known and straightforward .
we will briefly summarize the boundary conditions and the numerical method . since the basic equation ( [ basiceq11 ] ) is a second order differential equation ,
two boundary conditions should be given for @xmath61 at the center and the surface of the star .
the boundary condition at the center is that the function @xmath61 should be regular near @xmath67 .
this condition can be written by power series as @xmath68 at the surface , the lagrangian perturbation of pressure should be vanished , @xmath69 this condition can be written explicitly as @xmath70 we can numerically integrate the equation ( [ basiceq11 ] ) from the center with the condition ( [ bounarycondition1 ] ) , and from the surface with ( [ bounarycondition2 ] ) to the appropriate interior point . the eigen - value @xmath62 can be obtained as the consistent frequency so as to match functions smoothly at the point .
we also numerically calculate the rotational correction of normal frequency @xmath59 , by @xmath62 and @xmath61 through the equation ( [ defsigma1 ] ) . in order to check our numerical code for the relativistic cowling approximation , we calculated the fundamental normal mode frequencies having harmonic index @xmath71 in the range @xmath72 for the non - rotating stellar model used by ipser and lindblom .
our results were in good agreement with their normal frequencies and the difference was of the order less than @xmath73 .
we have calculated normal frequencies for a range of the stellar models with a polytropic equation of state ( [ eos ] ) .
we adapt two different polytropic indices ( @xmath74 ) .
the compactness of the star ranges from @xmath75 to @xmath76 the oscillation frequencies and the rotational corrections are calculated for the f- , p@xmath77- , and p@xmath78-modes having harmonics index @xmath79 in the range @xmath80 the results for typical stellar models are shown in tables 14 .
normal frequencies of the non - rotating stars and the rotational corrections are compared . since there is not so much difference in the polytropic index ,
we hereafter concentrate in the model with @xmath81 .
we show the normal frequencies and rotational corrections of the f - modes as a function of @xmath82 in fig.1 .
the eigen - frequencies of the non - rotating star , @xmath83 are given in fig.1(a ) , and the rotational corrections @xmath84 are in fig.1(b ) .
the results with the exact relativistic calculation are represented by the dashed line , and those with the cowling approximation are by the solid lines . figure 1(a ) shows that two results of the f - mode approach with the increase of @xmath82 , that is , the error monotonically decreases .
for example , the relative error for @xmath85 f - mode in @xmath86 model is about @xmath87 in the magnitude , but it decreases to @xmath88 in @xmath89 model .
the relativistic cowling approximation therefore gives better result in the f - mode calculation , as equilibrium model becomes more relativistic .
figure 1(b ) shows that both results are always in good agreement on the rotational corrections .
the relative error is always less than @xmath90 .
next , we show results of the p@xmath77- and p@xmath78-modes in figs.2 and 3 , respectively . compared with the exact calculation ,
the dependence of @xmath82 is similar in the oscillation frequencies and the rotational corrections .
the relative error is almost constant in @xmath83 , but increases in @xmath84 with @xmath82 .
the approximation for these p - modes becomes worse for the relativistic stars .
however , the discrepancy is always within @xmath91 error . from figs .
13 , it is quite clear that the accuracy of the cowling approximation increases with the spherical harmonic index , @xmath71 .
this feature is well - known in the newtonian cowling approximation .
( see , e.g. cox . )
the neglect of the gravitational perturbation is justified due to the averaging effect of the local fluctuations for high - order modes .
the newtonian cowling approximation also becomes better with the increase of the radial node number .
( see , e.g. , robe for numerical results of the newtonian cowling approximation . ) in the relativistic case , the accuracy also increases , but the behavior is a little complicated due to the additional parameter @xmath82 .
in this paper , we have compared the relativistic cowling approximation with the perturbation calculation of general relativity .
the accuracy of the cowling approximation is considerably good , concerning the normal frequencies in the non - rotating star and the first order rotational corrections .
remarkable fact is that the relative error of the f - mode decreases with the relativistic factor , @xmath82 .
the approximation is also better for larger radial node number and harmonic index @xmath92 , and greater central condensation , as in the newtonian pulsation theory .
one of the important applications of the normal modes is to determine the critical angular velocity of the rotating stars .
the maximum angular velocity is limited by the existence of the axisymmetric equilibrium state .
the equilibrium state however suffers secular instability due to viscosity and/or gravitational radiation reaction ( grr ) .
the instability sets in through @xmath93 ( viscous instability ) or @xmath94 ( grr instability ) , respectively ( e.g. , friedman and schutz ) .
form the newtonian calculation , most significant contributions to the grr instability are the f - modes with @xmath95 .
( see , e.g. , yoshida and eriguchi . )
the present study shows that the relativistic cowling approximation is able to give a good prediction of the oscillation frequency within @xmath96 error for the f - modes with @xmath97 in the typical neutron stars , @xmath98 .
the critical point is however not within the regime of the linear extrapolation from the non - rotating case , as suggested in the newtonian calculation .
that is , the present results , @xmath99 and @xmath100 are not sufficient , and the higher order rotational corrections are necessary .
therefore , solving normal frequency in rapidly rotating stars with the cowling approximation would be a good method to determine the critical angular velocity .
this work was supported in part by the grant - in - aid for scientific research fund of the ministry of education , science and culture of japan ( 08640378 ) .
s.y . would like to thank prof .
y. eriguchi and prof .
t. futamase for their continuous encouragement .
fig.1(a ) .
normal frequencies , @xmath83 vs. relativistic factor @xmath82 for f - mode and @xmath101 . dashed
lines and solid lines correspond to results with the relativistic perturbation calculation and the cowling approximation , respectively .
normal frequencies are normalized by @xmath104 . attached labels @xmath85 , @xmath105 , @xmath106 and @xmath107 mean the harmonic indices .
fig.1(b ) .
rotational corrections , @xmath84 vs. relativistic factor @xmath82 for f - mode and @xmath101 . dashed
lines and solid lines correspond to results with the full calculation and the cowling approximation , respectively .
rotational corrections are normalized by @xmath104
. attached labels @xmath85 , @xmath105 , @xmath106 and @xmath107 mean the harmonic indices . | we have calculated the non - radial oscillation in slowly rotating relativistic stars with the cowling approximation .
the frequencies are compared with those based on the complete linearized equations of general relativity .
it is found that the results with the approximation differ by less than about @xmath0 for typical relativistic stellar models .
the approximation is more accurate for higher - order modes as in the newtonian case .
oscillation , neutron star |
strong attraction between particles in solution leads to aggregation .
the kinetics of this aggregation process depends on the probability that a bond is formed when two particles collide .
two limiting cases are diffusion limited cluster aggregation ( dlca ) if bonds are formed at each collision and reaction limited cluster aggregation ( rlca ) if the probability to form a bond is very small . irreversible aggregation has been studied in detail both experimentally @xcite and using computer simulations @xcite .
the structure of clusters formed by random aggregation is self similar and characterized by a fractal dimension ( @xmath0 ) , which relates the radius of gyration ( @xmath1 ) to the aggregation number ( @xmath2 ) : @xmath3 .
the number of clusters with aggregation number @xmath2 can be described by a power law : @xmath4 .
as long as the clusters are on average far apart ( flocculation ) one finds @xmath5 and @xmath6 for dlca and @xmath7 and @xmath8 for rlca @xcite . with time
the clusters grow and the cumulated volume occupied by the clusters ( @xmath9 ) increases so that the average free space between the aggregates decreases .
when @xmath10 approaches the volume of the system the aggregates start to interpenetrate .
the aggregation process of highly interpenetrated clusters can be described by the percolation model and leads to gelation . for percolating clusters @xmath11 and @xmath12 @xcite .
the cross - over between flocculation and percolation occurs at a characteristic aggregation number ( @xmath13 ) and radius of gyration ( @xmath14 ) , that decrease with increasing particle volume fraction ( @xmath15 ) @xcite .
computer simulations of irreversible dlca and rlca have been done so far for hard spheres that form rigid bonds at contact @xcite . in this case
the aggregated particles are on average bound to two other particles , because ternary collisions are not possible .
however , in reality the bonds may be flexible , i.e. they may freely rotate .
one example is the much studied aggregation of spheres in the presence of other smaller particles through a depletion interaction @xcite .
another example is the aggregation of emulsion droplets with a slippery layer @xcite .
the latter experiment has motivated computer simulations of irreversible diffusion limited aggregation ( dla ) with flexible bonds @xcite .
the difference between dla and dlca is that during the former individual particles are allowed to diffuse until they collide with a single cluster @xcite , while during the latter all particles in the systems diffuse and collide to form many clusters .
dla leads to a self similar cluster with @xmath11 . in ref .
@xcite the dla simulation was modified to include free diffusion of the particle on the surface of the particle to which it is bound and was called slippery dla .
real random aggregation processes are , of course , better described by dlca . here
we report on a simulation study of dlca with finite interaction range in which the relative motion of bound particles is unhindered as long they remain within each others range . if the interaction range is very small compared to the radius of the particles this method could be called slippery dlca , but here we will use the expression flexible dlca for all interaction ranges .
flexible dlca ( or rlca ) should represent realistically the experimental systems mentioned above in the limiting case that the attraction energy is much larger than the thermal energy .
the results of flexible dlca will be compared to those obtained by dlca with rigid bonds .
the simulation method used here is called brownian cluster dynamics ( bcd ) .
a detailed description of the method and a comparison with molecular dynamics were reported elsewhere @xcite .
briefly , clusters are formed by connecting spheres within each others interaction range with probability @xmath16 .
particles are chosen randomly and moved a step size @xmath17 in a random direction unless it leads to overlap or breaks a bond .
the centre of mass displacement of the clusters is calculated and the clusters are moved cooperatively in the same direction so that the total displacement is inversely proportional to their radius , unless it leads to overlap .
bcd is equivalent to molecular dynamics if the cooperative cluster displacement is omitted as long as @xmath17 is sufficiently small .
systems with rigid bonds are simulated by performing only the cooperative cluster movements and not the individual particle displacements within the clusters .
the equilibrium state obtained by molecular dynamics and bcd is the same , but the dynamics depend strongly on whether the bonds are rigid or flexible .
irreversible dlca is simulated by setting @xmath18 .
the simulation is started with @xmath19 randomly distributed spheres with unit diameter in a box of size @xmath20 so that @xmath21
. the unit of time is set equal to the time needed for an isolated sphere to diffuse a distance equal to its diameter .
e.g. for spheres with diameter @xmath22 in water at @xmath23 the time unit is @xmath24 seconds .
the box size was varied up to @xmath25 and all the results shown here were not influenced by finite size effects unless specified .
in the following we show mainly results obtained with the interaction range fixed at @xmath26 , but we will briefly discuss the effect of varying the range .
it is well known that in the flocculation regime , i.e. @xmath27 the cluster growth during dlca can be described by the kinetic equations introduced by smolechowski @xcite : @xmath28 where @xmath29 is the so - called kernel that expresses the rate constant at which a cluster with aggregation number @xmath30 collides with a cluster with aggregation number @xmath31 : @xmath32 with @xmath33 the diffusion coefficient of the clusters and @xmath34 their collision radius .
@xmath33 is inversely proportional to the hydrodynamic radius ( @xmath35 ) of the clusters . for large clusters
both @xmath34 and @xmath35 are proportional to @xmath1 @xcite .
in fact , the growth kinetics of dlca can be well described by eq([e.1 ] ) using a constant kernel equal to @xmath36 .
it follows that the time dependence of the weight ( @xmath37 ) averaged aggregation number can be written as : @xmath38 in our simulations the particle diameter is unity and the diffusion coefficient of the individual spheres is @xmath39 so that @xmath40 . the ratio of the weight and the number ( @xmath41 ) averaged aggregation numbers becomes two for large @xmath37 .
these results have been confirmed using bcd with rigid bonds , but only if @xmath15 was very low . in @xcite
the interaction range was zero .
increasing the interaction range increases the collision radius and therefore @xmath36 , but the effect is small for @xmath26 .
for @xmath42 it is difficult to observe the flocculation regime because the crossover to the percolation regime occurs already when @xmath37 is still relatively small .
the growth of @xmath37 accelerates when @xmath10 approaches @xmath43 and @xmath37 diverges at the gel time ( @xmath44 ) .
this change of the growth kinetics between flocculation at short times and percolation close to @xmath44 has also been observed experimentally @xcite .
figure([f.1]a ) compares the increase of @xmath37 as a function of time for dlca with rigid and with flexible bonds at different volume fractions , while fig.([f.1]b ) shows the evolution of the polydispersity index ( @xmath45 ) .
the size of the clusters at the start of the simulation depends on the interaction range and the volume fraction , because more particles are within each others range if @xmath46 or @xmath15 are larger . for large @xmath46 and
@xmath15 the system percolates immediately , i.e. when the spheres are still randomly distributed , as was discussed in detail in @xcite .
the percolation threshold decreases with increasing @xmath15 and is @xmath47 for @xmath26 . at @xmath48 the increase of @xmath37 with time
was linear and @xmath45 was close to two as expected for the flocculation regime .
the results were similar for rigid and flexible dlca , but rigid dlca showed the influence of a transition to the percolation regime sooner .
the gel point was not reached for this volume fraction during the simulation which took several weeks of computer time . for @xmath49 the flocculation regime
was not observed and the growth of both the size and the polydispersity of the clusters accelerated until they diverged at @xmath44 .
the growth of the clusters and gelation was somewhat faster for flexible dlca . at @xmath50
the increase of @xmath37 was initially slightly faster during flexible dlca , but then the growth became almost linear over a short period of time while the polydispersity index remained constant .
after this period the increase of @xmath37 and @xmath45 accelerated until the gel was formed . during rigid dlca the slowing down of the aggregation was not observed so that gelation at this volume fraction was faster than for flexible dlca .
the growth rate at different volume fractions can be understood by considering the structure of the clusters as will be discussed below .
figure([f.2 ] ) shows @xmath51 for @xmath52 obtained at different volume fractions . at the lowest concentration we observed a bell shape as is also observed for rigid dlca in the flocculation regime @xcite . at higher volume fractions
@xmath51 followed approximately the power law decay expected for the percolation regime : @xmath53 .
we note that the cut - off function of the size distribution of percolating clusters has a gaussian shape @xcite .
more extensive simulations on even larger boxes will be needed to test whether @xmath51 obtained from flexible dlca has exactly the same functional form as for rigid dlca or static percolation .
aggregation leads to an increase of the number of bonds per particle ( @xmath54 ) . in fig.([f.3]a )
the average number of bonds per particle ( @xmath55 ) is plotted as a function of time . as mentioned above , the starting value of @xmath55 is larger for larger @xmath15 because more randomly distributed spheres are within each others interaction range . at low volume fractions ,
@xmath55 increased during rigid dlca until it stagnated at a value close to @xmath56 , while at large @xmath15 the increase of @xmath55 was small because almost all the particles were already part of the final structure at the start of the aggregation . during flexible dlca @xmath55
increased sharply at first , but when it had reached a value around @xmath57 further growth of @xmath55 became very slow .
contrary to the case of rigid dlca , @xmath55 continued to increase over the whole duration of the simulation even after all spheres were attached to the network indicating that restructuring of the system occurred .
this process was very slow and may be called ageing . for @xmath49
most of the restructuring happened after the gel point and continued even when the sol fraction had become very small . in fig.([f.3]b )
are plotted the values of @xmath55 at the end of the simulation as a function of the volume fraction .
@xmath55 has a minimum at @xmath58 , the origin of which will be discussed below .
the distribution of @xmath54 at long times only weakly depended on the volume fraction for @xmath59 .
it was approximately gaussian with a half width of @xmath60 .
it started at @xmath61 and peaked close to @xmath55 .
very few particles had @xmath62 showing that no crystallisation had occurred . at higher volume fractions
the distribution shifted to larger @xmath54 . during rigid dlca
the cluster configuration is determined by random collisions and the clusters can not rearrange once they are formed . therefore rigid dlca yields in the flocculation regime clusters with the same average radius of gyration for a given aggregation number independent of time .
this is not the case for flexible dlca where the particles in the clusters rearrange until a maximum of bonds are formed without breaking any existing bond . in this way
the density of the clusters increases with time . at very low volume fractions
the growth rate was slower than the restructuring time so that at each moment the cluster configuration had reached the steady state , while at high volume fractions the gel was formed quicker than the time needed to restructure .
therefore the value of @xmath55 at the gel point was close to the final value for @xmath63 , but much lower at large @xmath15 , see fig.([f.3]a ) .
figure([f.4]a ) shows images of clusters formed during flexible dlca at very low volume fractions .
the configuration of clusters up to @xmath64 was unique contrary to clusters formed by rigid dlca . for @xmath65 one configuration
was observed in @xmath66 of the cases , but one other configuration was also possible . with increasing aggregation number more different configurations were found .
large clusters were generally elongated strands that branched when @xmath2 exceeded about @xmath67 .
the randomly branched structures resembled clusters formed by rigid dlca , except that the strands were much thicker than the single particle diameter , see fig.([f.4]b ) .
the basic unit of all larger clusters formed by flexible dlca was a tetrahedron and for @xmath68 particles with less than @xmath69 neighbours were not observed .
the fact that the tetrahedron is the basic structural unit does not imply however that tetrahedra were formed first and subsequently aggregated to form larger clusters as was suggested in @xcite . at all stages of the aggregation we observed all aggregation numbers with no preference for multiples of @xmath60 . as mentioned in the introduction ,
the large scale structure of clusters formed by rigid dlca in the flocculation regime is self similar with @xmath5 . in fig.([f.5]a )
a comparison is shown of the dependence of @xmath2 on @xmath1 between clusters formed by flexible dlca and by rigid dlca at @xmath48 . in both cases @xmath70 , but the prefactor was larger for flexible dlca ( @xmath71 ) than for rigid dlca ( @xmath72 ) because the local structure of clusters is denser .
figure([f.5]b ) shows the dependence of @xmath2 on @xmath1 for clusters formed at @xmath73 close to the gel point .
the structure of the clusters was almost the same for flexible and rigid dlca because not much restructuring had yet occurred before the gel was formed . for larger clusters @xmath74 as predicted by the percolation theory . at intermediate volume fractions
the density of the clusters increased with time for flexible dlca and the transition between flocculation and percolation shifted to larger values of @xmath2 with decreasing @xmath15 .
the structure of the clusters at higher volume fractions was the same for rigid and flexible dlca , yet the growth rate was faster for flexible dlca .
this can be explained by the fact that neighbouring fractal clusters have a larger probability to bind when they are flexible than when they are rigid .
it is likely that in the case of rigid clusters rotation would also accelerate the growth during the percolation regime . at low volume fractions both the collision radius and
the hydrodynamic radius of clusters formed during flexible dlca were smaller for a given aggregation number .
apparently , both effects compensated to give similar growth kinetics for flexible and rigid dlca .
however , the percolation regime was reached at larger @xmath2 during flexible dlca than during rigid dlca .
densification of the clusters during flexible dlca explains why the growth rate at @xmath50 was initially faster and at later times slower than for rigid dlca .
initially , the growth was faster due to flexibility , but later densification rendered the clusters smaller so that the transition to the percolation regime was delayed .
the gel structure can be characterized by the pair correlation function ( @xmath75 ) or its fourier transform the static structure factor ( @xmath76 ) .
@xmath75 represents the average number concentration of particles at a distance @xmath77 from any given particle and reaches the number concentration of the system @xmath78 at large @xmath77 .
an extensive study of gels formed by rigid dlca was reported in @xcite .
it was shown that pair correlation functions of gels at low volume fractions had distinct features at small @xmath77 followed by a power law decay ( @xmath79 with @xmath5 ) for @xmath80 .
@xmath75 had a weak minimum at a characteristic value @xmath81 close to the correlation length of the concentration fluctuations before it reached @xmath82 .
the correlation length decreased with increasing volume fraction , and became of the order of a few particle diameters for @xmath83 . as a consequence ,
the fractal structure did not exist at higher volume fractions .
@xmath75 showed a delta peak at @xmath84 representing the contribution of the two nearest neighbours and increased continuously to @xmath85 at @xmath86 starting from a low value close to @xmath84 . at low volume fractions the main contribution to @xmath75 in this range came from the bound next - nearest neighbours .
their contribution stopped at @xmath86 which caused a discontinuous drop of @xmath75 .
a small inflection of @xmath75 was observed at @xmath87 marking the influence of the second shell and at larger @xmath77 the power law decay started .
the local structure of the gels was identical for small volume fractions ( @xmath63 ) but changed at higher volume fractions because positional correlations between randomly distributed spheres at @xmath88 were no longer negligible .
figure.([f.6]a ) shows pair correlation functions of gels obtained by flexible dlca at different volume fractions .
the values of @xmath75 within the interaction range were much larger than for @xmath89 and are shown separately in fig.([f.6]b ) .
there is a discontinuity in @xmath75 at @xmath90 , i.e. the maximum distance between a pair of bonded particles .
this discontinuity is a consequence of the square well interaction and is not seen for a continuous potential such as the lennart - jones potential . for @xmath89
, peaks can be seen indicating a high degree of local order that was absent for rigid dlca .
the maxima at @xmath91 and @xmath92 are characteristic for the tetrahedral structure with a bond length between @xmath93 and @xmath94 . if this structure is extended in a regular linear fashion it leads to the so - called bernal spiral .
small peaks at larger distances indicate that the order persists to some extent but for @xmath95 the structure became self similar . at even larger distances a very weak minimum
was observed at small volume fractions similar to that found for rigid dlca , but at a given volume fraction it was situated at larger @xmath77 values for flexible dlca . at a given volume fraction , the correlation length of the gel is larger for flexible dlca because the structure is locally denser . for the same reason the radius @xmath14 at the crossover between flocculation and percolation was larger .
it was shown in @xcite for rigid dlca , that the correlation length and @xmath14 are close .
bonded nearest neighbours can be situated at any distance within the interaction range , i.e. between @xmath93 and @xmath94 in the present case , but fig([f.6]b ) shows that the distance was not uniformly distributed .
in fact there was a preference for distances close to @xmath93 or @xmath96 , which means that some bonds were compressed and others stretched .
the local structure was independent of the volume fraction for @xmath97 , because it is mainly determined by the aggregation process of particles that were initially outside each others range .
at higher volume fractions a significant number of particles were in contact before the aggregation process started .
the restructuring was thus more constrained and leading to less local order .
the split between the peaks at @xmath98 and @xmath56 became less distinct and resembled more closely the split peak observed for super cooled liquids .
the corresponding structure factors are shown in fig.([f.7 ] ) for different volume fractions . at small @xmath99 a maximum
was found at a value ( @xmath100 ) that increased with increasing concentration .
for rigid dlca it was shown that the position of maximum is inversely proportional to that of the minimum of the pair correlation function @xmath101 @xcite . over a narrow @xmath99-range @xmath76 decreased with increasing @xmath99 following a power law , which is expected for self similar structures : @xmath102 .
the data are compatible with @xmath5 , see the solid line in fig.([f.7 ] ) , but for a precise determination of the fractal dimension even smaller volume fractions need to be investigated . at large @xmath99-values
@xmath76 oscillated around unity .
figure([f.8 ] ) compares the structure factors of stable gels formed by flexible and rigid dlca at @xmath50 .
the maximum has a much larger amplitude and is situated at smaller @xmath99-values for gels formed by flexible dlca and the oscillations at large @xmath99 are stronger .
the reason is that the local density of the gels is higher and therefore the correlation length is larger . at this volume fraction
a large scale fractal structure can be observed if the gels are produced by flexible dlca , but not if they are produced by rigid dlca .
the evolution of @xmath75 and @xmath76 as a function of time during rigid dlca was shown in @xcite . with time
the maximum of @xmath76 shifted to lower @xmath99 and increased in amplitude , while the minimum of @xmath75 shifted to larger @xmath77 , because the correlation length increased . during rigid dlca
the structure is fixed as soon as the bonds are formed , but this is not the case for flexible dlca .
we illustrate this point by showing the evolution of the structure at @xmath103 during flexible dlca . at this volume fraction
the local structure is not yet influenced by correlation between the randomly distributed spheres .
@xmath75 is plotted in fig([f.9 ] ) at different times . at the start of the aggregation , @xmath75 was constant within the bond range , i.e. @xmath104 , and equal to @xmath82 .
with time the number of bonds increased , but @xmath75 remained constant in the range even when the gel was formed . at longer times , however , @xmath75 became larger at @xmath84 and @xmath105 , implying that in order to maximize the number of bonds under constraint , many bonds needed to be either stretched or compressed during restructuring of the gel .
the peaks at @xmath98 and @xmath86 appeared already before the gel point indicating that some tetrahedral structure had been formed without bond stretching or compression . the characteristic minimum of @xmath75 at @xmath81 developed slowly and moved to longer distances .
figure([f.10 ] ) shows the corresponding evolution of the structure factor .
the shift of the maximum to lower @xmath99 and the increase of its amplitude are caused by an increase of the correlation length .
the amplitude of the oscillations at high @xmath99-values increased due to local densification . at low volume fractions the pair correlation function and the structure factor changed very little after the gel point because the aggregation was sufficiently slow so that the clusters had time to restructure before they percolated . at high volume fractions gels were formed very rapidly and the structure at the gel point resembled that of gels formed by rigid dlca . in this case , the distinguishing features of the flexible dlca gels developed after the gel point . at long times the structural changes became increasingly slow , but they did not stop completely during the simulation time . for rigid dlca the mean squared displacement ( msd ) of the spheres in the gels is , of course , zero .
however , in flexible dlca gels the spheres have a significant mobility , which can be characterised by measuring the msd as a function of time .
it is important to distinguish displacements due to restructuring from displacements due to flexibility .
the displacement due to restructuring can be probed by measuring the average total msd from the start of the aggregation process .
the total msd as a function of time increases rapidly at first followed by a very weak increase when all particles are part of the gel .
one can characterise the flexibility of the gels by measuring the msd over a period of time during which restructuring is negligible , i.e. for gels that have aged for a long time .
alternatively , one can stop further bond formation and thus further restructuring .
this is , of course , easier to do in simulations than in experiments on real systems .
figure([f.11]a ) shows the msd of spheres in gels at the latest simulation time , when all particles were part of the percolating network .
initially , free diffusion was observed over very short distances .
then the average msd slowed down and finally stagnated at a value @xmath106 .
@xmath107 increased with decreasing concentration , but we could not properly observe @xmath107 at low volume fractions because the times scales involved were too large .
if the displacement of the particles is caused by brownian motion , the distribution of @xmath108 ( or the self part of the van hove correlation function @xmath109 ) can be described by a gaussian function .
figure([f.11]b ) shows @xmath109 when the average was close to @xmath107 .
it is clear that the displacement was highly heterogeneous as might be expected from the heterogeneous structure of the gels .
the non - gaussian character can be expressed by the parameter @xmath110 which is zero for a gaussian function .
we found @xmath111 for all volume fractions .
the self part of the intermediate scattering function was found to fully decay at scattering wave vectors @xmath112 , while it showed a plateau at long delay times for smaller @xmath99 at a value that increased with decreasing @xmath99 .
this is , of course , a direct consequence of the restricted msd of the particles , see below . for rigid dlca
the effect of increasing the interaction range is twofold . in the first place ,
as mentioned above , the collision radius of the spheres is increased by a factor @xmath113 . as a consequence
the clusters are larger and the aggregation is faster . in the second place ,
the concentration of bonds between the randomly distributed spheres at the start of the aggregation is larger .
the latter effect becomes important when the average distance between nearest neighbours ( @xmath114 ) is smaller than @xmath113 ( @xmath115 for @xmath116 ) .
the percolation threshold of randomly distributed spheres decreases with increasing @xmath46 was discussed in @xcite . however , as long as @xmath117 is much larger than @xmath113 , i.e. at small volume fractions , the effect of the interaction range on the structure of the clusters and the gels is small and the fractal dimension remains the same . for flexible dlca
the same effects are present , but in addition the system has a larger degree of freedom to increase the number of bonds by restructuring if @xmath46 is larger . as a consequence the value of @xmath55 of the gels increases with increasing @xmath46 .
for @xmath118 we found @xmath119 after long times at low volume fractions .
a qualitatively different local structure is formed when @xmath46 is larger than @xmath120 because the tetrahedral configuration is no longer the basic unit .
consequently all spheres are bound to all other spheres for pentamers and hexamers .
figure([f.12 ] ) shows a comparison of the structure factor at @xmath50 for @xmath26 and @xmath118 .
the maximum is situated at larger @xmath99 for @xmath118 and has a smaller amplitude , indicating that the correlation length is smaller if @xmath46 is larger .
the second maximum is situated at smaller @xmath99 indicating that the local structure ( the strand thickness ) is larger and the oscillations at high @xmath99 have a lower amplitude indicating less order .
the comparison shows that when the interaction range is larger then the gels are more homogeneous with thicker strands and less local order .
the effect of the interaction range on the structure is small for small @xmath46 and we expect that using a smaller interaction range than @xmath121 will have little influence on the structure .
experimentally , structures similar to those found in the present simulations of flexible dlca have been reported for mixtures of hard spheres and polymers in which an effective attraction between the spheres is induced by depletion of the polymers @xcite .
if the concentration of polymers is high the attraction may become so strong that the aggregation is irreversible on the time scale of observation . in this case no significant evolution of the structure is observed after gelation .
we note that in experiments matters are sometimes complicated by electrostatic interaction .
recently , experiments on aggregating emulsion droplets were reported where the bonds were claimed to be truly irreversible , but bound droplets were free to move along the interface @xcite .
the structure factor obtained from the experiments was compared to results of flexible dla simulations @xcite .
many of the features that distinguish the structure of clusters formed by rigid dlca from those formed by flexible dlca have also been observed in simulations of rigid and flexible dla with zero interaction range ( note that the expressions classic and slippery were used in @xcite to refer to rigid and flexible , respectively ) .
also in dla , flexible bonds caused the formation of thicker strands with a local tetrahedral configuration .
as for dlca , the fractal dimension of dla clusters was the same for flexible and rigid bonds .
the distribution of z was similar for flexible dla and dlca , and the structure factor showed prominent oscillations at high @xmath99-values in both cases . in fig.([f.13 ] ) the structure factor obtained from flexible dlca ( @xmath26 ) is compared to that obtained from flexible dla ( @xmath122 ) and experiments .
the experimental structure factor was obtained by dividing the scattering data of the system with the particle form factor of the droplets .
the latter was obtained experimentally before the aggregation had started and was assumed to be not influenced by interaction . as mentioned above ,
the aggregation process is more realistically described by dlca , which results in a smaller fractal dimension ( @xmath5 ) than obtained by dla ( @xmath11 ) .
nevertheless , the results from the two types of simulations are remarkably close at large @xmath99-values , indicating that almost the same local structure is formed .
this is perhaps not surprising if one considers that local restructuring occurred in both cases until the maximum number of bonds was formed under almost the same constraints .
the simulations reproduced the position of the peak at @xmath123 seen in the experiments , but the amplitude was somewhat smaller ( notice that the experimental data have been shifted upward for the comparison with the simulation results in figure 4 of @xcite ) .
the upturn at low @xmath99-values started at larger @xmath99 in the simulations than in the experiments .
however , one should be careful when drawing conclusions on the basis of the deviation at low @xmath99-values , because the experimental results are very sensitive to the exact shape of the particle form factor in this @xmath99-range .
as far as we are aware , flexible dlca with strictly irreversible bonds has not been simulated before , but the present results may be compared with molecular dynamics simulations of reversible dlca in the limit of very strong attraction .
foffi et al .
@xcite studied the structure and dynamics of spheres with a square well interaction with width @xmath124 and two different well depths : @xmath125 and @xmath126 @xmath127 .
the interaction strengths may also be expressed in terms of the second virial coefficient @xcite : @xmath128 and @xmath129 , in units of the particle volume . at both interaction
strengths the systems are far below the binodal of the liquid - liquid phase separation .
crystallisation was avoided by using a bidisperse distribution of spheres with slightly different sizes . at smaller interaction strength
the system slowly coarsened with time , but at @xmath130 the bonds may be considered irreversible on the time scale of the simulation .
this situation is thus comparable to flexible dlca .
the features of the gels obtained by foffi et al .
were very close to the ones reported here .
unfortunately , the authors did not show pair correlation functions , which would allow a more detailed comparison of the local structure .
but the images of the clusters and the static structure factors of the gels resemble closely those shown here .
the distribution of the bond coordination number was almost the same as the one found in this study , but @xmath55 was a bit smaller which can be explained by the narrower interaction range .
foffi et al . also observed a minimum of @xmath55 at volume fractions close to @xmath131 , which they speculated to be related to the critical point of the liquid - liquid phase separation .
we believe that the minimum is caused by a combination of opposing effects . at very low volume fractions
the clusters can restructure while they grow , but with increasing volume fraction the growth rate increases and much of the restructuring has to occur after the percolation threshold .
this increases the constraints on the restructuring and thus lowers @xmath55 with increasing @xmath15 . at high volume fractions
, however , the number of bonds increases due to crowding even for randomly distributed spheres .
this leads to an increase of @xmath55 with increasing @xmath15 .
the latter effect is more important when @xmath46 is larger , which explains the stronger increase of @xmath55 found in the present study .
foffi et al . argued that the gel was formed by _ a phase separation process interrupted by attractive glass transition_.
the same idea has also been put forward by others @xcite .
it is clear , that gelation occurs only for interaction strengths where the equilibrium state would be phase separated and therefore one might indeed call the gel formation an interrupted phase separation
. however , the suggested origin of the interruption is debatable .
the concept of attractive glass was introduced to account for the slowing down of the dynamics at high concentrations with increasing attraction . even if attractive glass formation would be a useful concept for hard spheres with a narrow interaction range at high volume fractions , it is not clear what it means in the context of the arrested state at low concentrations .
many particles are situated at the surface of the network strands and can not be in a glassy state .
these surface particles can escape from their neighbours by breaking all the bonds .
this process inevitably leads to coarsening .
the arrest is only inferred if the observation time is shorter than the time needed for surface particles to escape .
we have done preliminary simulations that showed that the rate of coarsening decreased with decreasing @xmath132 following a power law and is simply too slow at @xmath129 to observe on the time scale probed by foffi et al .. the results presented here may also be compared with earlier brownian dynamics simulations of colloidal gels formed by monodisperse spheres by darjuzon et al.@xcite . in those simulations the spheres interacted with a continuous potential that was steeply repulsive at contact and attractive over a range of @xmath121 .
the shape of the potential was chosen to be close to that of depletion interaction . above a certain interaction strength the spheres crystallised @xcite , but when the interaction was very strong crystallisation was no longer observed for the duration of the simulation .
a detailed study was done of the dynamics of this system at a single volume fraction @xmath133 and a minimum interaction potential of @xmath134 . for this interaction strength the life time of the bonds near
the potential minimum is very long compared to the duration of the simulation .
however , stretched bonds were still rapidly broken and reformed , because the potential goes to zero at the outer limit of the interaction range . in spite of the different shape of the potential and the reversibility of stretched bonds many features of this simulation
are similar to those reported here for strictly irreversible aggregation with the same bond range .
it was also found that @xmath55 increased rapidly at first followed by a very slow increase .
a similar distribution of the bond coordination number was found at long times and @xmath75 showed similar features notably the two peaks indicating tetrahedral structure .
the msd of spheres in the gel and also the distribution of displacements at @xmath133 were similar to those reported here .
the authors argued that displacements larger than one particle diameter were due to particles breaking of the network .
it is clear from the present simulations , however , that even in irreversibly bound gels large scale mobility is possible due to cooperative motion .
it is probable that at lower volume fractions the maximum displacement is determined by the correlation length of the gels .
we note that in a recent simulation study of more open gels formed by restricting the bond angles and thus avoiding the formation of dense strands , it was also noted that the msd displacement can be quite large due to cooperative motion @xcite .
darjuzon et al . studied the self - intermediate scattering function in detail and found a fast decay towards a plateau followed by a slow decay to zero .
the value of the plateau was shown to be directly related to @xmath135 .
the relaxation time of the slow decay increased with increasing waiting time and became very slow .
preliminary calculations showed similar behaviour for the square well system studied here .
the similarities between the two systems indicate that the reversibility of stretched bonds is not an essential feature .
we believe that in both systems the aging is caused by the formation of more ( strong ) bonds under the constraint that existing ( strong ) bonds can not break at least for the duration of the simulation .
lodge and heyes @xcite studied spheres interacting with a lennart - jones potential with varying interaction range and depth using brownian dynamics simulations . a similar shape of the pair correlation function was found if the range was narrow and the interaction strong .
however , they also observed a peak at a smaller distance indicating that some crystallisation had occurred .
the authors studied the phase separation kinetics up to @xmath136 in the time units used here .
the kinetics was very slow in the case of strong attraction and may thus also be called aging . in fact the time dependence of the peak position shown in fig.([f.10 ] ) is close to that obtained by lodge and heyes using a lennart - jones potential with a narrow interaction range .
this indicates again that when the interaction is strong , the breaking of bonds is not important for the aging process at least in the early stages .
irreversible dlca with flexible bonds causes locally densification , but on large length scales the structure is the same as for dlca with rigid bonds .
the fractal dimension of clusters formed in very dilute systems is @xmath137 in both cases .
locally the systems have a tetrahedral structure and show a certain degree of order that is independent of the volume fraction for @xmath97 . at higher volume fraction the order is less distinct and similar to that of super cooled liquids . the system tries to maximize the number of bonds under the constraint of no bond breaking .
the increase of the bond coordination number is fast at first , but it becomes progressively slower .
the slow restructuring ( aging ) of the gels continues for very long times . at low volume fractions
most of the local densification occurs while the clusters are formed , but at high volume fractions ( @xmath138 ) the restructuring occurs mainly after gelation . at very low volume fractions
the growth rate of the clusters during flexible dlca is the same as for rigid dlca and can be understood in terms of smolechowski s kinetic equations . at high volume fractions
the growth rate is somewhat faster , because flexibility increases the collision rate .
gels formed at low volume fractions show a large degree of flexibility .
the average msd stagnates at a value that increases with decreasing volume fraction . however , the msd of the particles is highly heterogeneous reflecting the fractal structure of the gels .
many features that appeared during irreversible flexible dlca closely resemble simulations and experiments reported in the literature on spheres with a strong but finite attractive interaction . for the interpretation of these studies
it is essential to consider whether the effect of bond breaking is significant during the time of observation . j. n. wilking , s. m. graves , c. b. chang , k. meleson , m. y. lin , and t. g. mason , phys .
lett . * 96 * , 015501 ( 2006 ) .
m. carpineti , f. ferri , m. giglio , e. paganini , and u. perini , phys .
a * 42 * , 7347 ( 1990 ) .
t. nicolai and s. cocard , eur .
j. e * 5 * , 221 ( 2001 ) .
d. a. weitz and j. s. huang , phys .
lett . * 54 * , 1416 ( 1985 ) .
m. y. lin , h. m. lindsay , d. a. weitz , r. klein , r. c. ball , and p. meakin , j. phys . : condens .
matter * 2 * , 3093 ( 1990 ) .
m. y. lin , h. m. lindsay , d. a. weitz , r. c. ball , r. klein , and p. meakin , phys .
a * 41 * , 2005 ( 1990 ) .
p. meakin , phys .
lett . * 51 * , 1119 ( 1983 ) .
m. kolb , r. botet , and r. jullien , phys .
lett . * 51 * , 1123 ( 1983 ) .
s. diez orrite , s. stoll , and p. schurtenberger , soft matter * 1 * , 364 ( 2005 ) .
j. c. gimel , d. durand , and t. nicolai , phys .
b * 51 * , 11348 ( 1995 ) .
j. c. gimel , t. nicolai , and d. durand , j. sol - gel sci .
* 15 * , 129 ( 1999 ) .
j. c. gimel , t. nicolai , d. durand , and j .-
m . teuler , eur .
j. b * 12 * , 91 ( 1999 ) .
m. rottereau , j. c. gimel , t. nicolai , and d. durand , eur .
j. e * 15 * , 133 ( 2004 ) .
m. rottereau , j. c. gimel , t. nicolai , and d. durand , eur .
j. e * 15 * , 141 ( 2004 ) .
c. r. seager and t. g. mason , phys .
e * 75 * , 011406 ( 2007 ) .
g. odriozola , a. moncho - jorda , a. schmitt , j. callejas - fernandez , r. martinez - garcia , and r. hidalgo - alvarez , europhys . lett . *
53 * , 797 ( 2001 ) .
p. meakin and f. family , phys .
a * 36 * , 5498 ( 1987 ) .
p. meakin , phys .
scr . * 46 * , 295 ( 1992 ) .
d. stauffer and a. aharony , _ introduction to percolation theory _
( taylor & francis , london , 1992 ) .
a. e. gonzalez , phys .
lett . * 71 * , 2248 ( 1993 ) .
a. d. dinsmore and d. a. weitz , j. phys .
: condens .
matter * 14 * , 7581 ( 2002 ) .
a. i. campbell , v. j. anderson , j. s. van duijneveldt , and p. bartlett , phys . rev
94 * , 208301 ( 2005 ) .
p. j. lu , j. c. conrad , h. m. wyss , a. b. schofield and d. a. weitz , phys .
96 * , 028306 ( 2006 ) . n. b. simeonova , r. p. a. dullens , d. g. a. l. aarts , v. w. a.
de villeneuve , h. n. w. lekkerkerker and w. k. kegel , phys .
e * 73 * , 041401 ( 2006 ) .
k. n. pham , s. u. egelhaaf , p. n. pusey , and w. c. k. poon , phys .
e * 69 * , 011503 ( 2004 ) .
s. manley , h. m. wyss , k. miyazaki , j. c. conrad , v. trappe , l. j. kaufman , d. r. reichman , and d. a. weitz , phys . rev .
lett . * 95 * , 238302 ( 2005 ) .
s. babu , j. c. gimel , t. nicolai and c. de michele arxiv : cond - matt/0711.0308v1 .
m. von smoluchowski , z. phys .
* 17 * , 585 ( 1916 ) .
m. von smoluchowski , z. phys . chem . *
92 * , 129 ( 1917 ) .
s. chandrasekhar , rev .
phys . * 15 * , 1 ( 1943 ) .
m. lattuada , h. wu , and m. morbidelli , j. colloid interface sci .
* 268 * , 96 ( 2003 ) .
h. wu , j. xie , and m. morbidelli , biomacromolecules * 6 * , 3189 ( 2005 ) .
m. rottereau , j. c. gimel , t. nicolai , and d. durand , eur .
j. e * 11 * , 61 ( 2003 ) .
g. odriozola , r. leone , a. schmitt , j. callejas - fernandez , r. martinez - garcia and r. hidalgo - alvarez , j. chem . phys . * 121 * , 5468 ( 2004 ) .
j. c. gimel , t. nicolai , and d. durand , j. phys .
a : math . gen . * 33 * , 7687 ( 2000 ) .
g. foffi , c. de michele , f. sciortino , and p. tartaglia , j. chem . phys . * 122 * , 224903 ( 2005 ) .
t. kihara , rev .
phys . * 25 * , 831 ( 1953 ) .
a. m. puertas , m. fuchs , and m. e. cates , phys .
lett . * 88 * , 098301 ( 2002 ) .
a. m. puertas , m. fuchs , and m. e. cates , phys . rev .
e * 67 * , 031406 ( 2003 ) .
k. a. dawson , g. foffi , m. fuchs , w. gtze , f. sciortino , m. sperl , p. tartaglia , t. voigtmann , and e. zaccarelli , phys .
e * 63 * , 011401 ( 2000 ) .
e. zaccarelli , f. sciortino , p. tartaglia , g. foffi , g. d. mccullagh , a. lawlor , and k. a. dawson , physica a * 314 * , 539 ( 2002 ) .
f. cardinaux , t. gibaud , a. stradner , and p. schurtenberger , phys .
99 * , 118301 ( 2007 ) .
r. j. m. d arjuzon , w. frith , and j. r. melrose , phys .
e * 67 * , 061404 ( 2003 ) .
k. g. soga , j. r. melrose , and r. c. ball , j. chem . phys . * 110 * , 2280 ( 1999 ) . e. del gado and w. kob , phys .
* 98 * , 028303 ( 2007 )
. j. f. m. lodge and d. m. heyes , phys .
* 1 * , 2119 ( 1999 ) . | irreversible diffusion limited cluster aggregation ( dlca ) of hard spheres was simulated using brownian cluster dynamics .
bound spheres were allowed to move freely within a specified range , but no bond breaking was allowed . the structure and size distribution of the clusters
was investigated before gelation . the pair correlation function and the static structure factor of the gels
were determined as a function of the volume fraction and time .
bond flexibility led to local densification of the clusters and the gels , with a certain degree of order . at low volume fractions densification of the clusters occurred during their growth , but at higher volume fractions it occurred mainly after gelation . at very low volume fractions ,
the large scale structure ( fractal dimension ) , size distribution and growth kinetics of the clusters was found to be close to that known for dlca with rigid bonds .
restructuring of the gels continued for long times , indicating that aging processes in systems with strong attraction do not necessarily involve bond breaking .
the mean square displacement of particles in the gels was determined .
it is shown to be highly heterogeneous and to increase with decreasing volume fraction . |
fluxon dynamics in long josephson junctions is a topic of strong interest due to its rich nonlinear properties and applications in fast electronics , in particular as a radiation source at high frequencies @xcite .
an extension of that system is to form a metamaterial by stacking several josephson junctions on top of each other , which are modeled by @xmath0 coupled partial differential equations .
such superconductors are employed in a variety of devices ( stacks , flux - flow devices , and synchronized arrays ) and are capable of power emission in the range 0.5 - 1 thz .
integration in arrays could give an improvement in the power performances , above @xmath1 @xcite .
practical applications are especially in the field of bio - sensing , nondestructive testing and high speed wireless communications @xcite . for such reasons
we aim to understand if some simple mechanism is at work in all these devices .
such a system is used as a model for high temperature superconductors of the bscco type @xcite . in this communication
we go one step further in complexity and include results on a nonlinear behavior of the shunt resistor , giving rise to features similar to stacked josephson junctions coupled to a cavity @xcite .
such a model is needed in order to understand and interpret the experimental measurements . for frequencies in the ghz or even thz range , either an intrinsic or an external cavity
is needed to enhance the radiated power to useful levels .
figure [ fig1]a shows qualitatively the appearance of a nonlinear current - voltage ( iv ) curve for the quasi particle tunneling in the josephson junction .
the particular form of the iv curve resembles a distorted n and hence we refer to this particular form as @xmath0-shaped iv curve .
note that the quasi particle tunnel current is a unique function of the applied voltage , but the inverse function is not unique .
similarly , fig .
[ fig1]b depicts an iv curve , which is shaped as a distorted @xmath2 ( @xmath2-shaped iv curve ) . in this latter case
the voltage is a unique function of the current . in general
the nonlinear behavior leading to a negative differential resistance ( ndr ) of josephson junctions plays a key role for applications .
an example is a parametric amplifier or a radiation source at high frequencies @xcite .
examples of ndr are numerous : ( i ) josephson junction with a cavity @xcite , ( ii ) backbending of the bscco energy gap @xcite , ( iii ) structure at the energy gap difference if the junction consists of two different superconductors , ( iv ) in connection with fiske steps , zero field steps @xcite and even in rf - induced steps @xcite . in some cases of ndr
a nonlinear region that can not be explained shows up in the iv curve @xcite .
the two qualitatively different cases of a nonlinear differential resistance , referred to as @xmath0-shaped and @xmath2-shaped regions , observed in experiments will be discussed in the next sections .
we mention that besides the ndr at finite voltages , also absolute negative resistance @xcite and negative input resistance at zero voltages @xcite have been reported . in this work
we want to emphasize the role of nonlinearities and to show that even a very simple model can give rise to an interesting power emission profile . however , our model , being a high frequency model , can not capture effects at low timescale , such as thermal effects @xcite .
we discuss below various examples of a negative differential resistance in josephson junctions .
josephson junctions come in a variety of forms with different properties but the same generic behavior .
some examples are : ( i ) the traditional low temperature ( @xmath3 ) josephson junction with qualitatively different behaviors depending on the dimensions of the junction , ( ii ) high @xmath3 intrinsic josephson junctions that are typically described as a stack of long josephson junctions leading to coupled sine - gordon equations and ( iii ) point contacts and microbridges that are the easiest to describe mathematically .
some features are generic , like the josephson voltage to frequency relation , the supercurrent , the energy gap etc . in all cases
we may have a coupling to a cavity either intrinsic ( internal ) or external , which of course complicates the mathematics but may be important for applications .
the two different cases of negative differential resistance @xmath0-shaped and @xmath2-shaped discussed here are shown in fig .
[ fig1 ] .
-shape , and b ) @xmath2-shape .
the lines are the iv curves ( showing hysteresis ) .
the dashed lines indicate the bias , voltage for the @xmath0-shape resistor , current for the @xmath2-shape case .
the short - dashed lines show the switching from a branch to another when the bias is reversed /current bias for the @xmath0-shape , voltage bias for the @xmath2-shape .
finally , @xmath4 and @xmath5 represent the dip of the negative differential resistor , see sections [ sshape ] and [ figure].,title="fig:",width=264 ] + -shape , and b ) @xmath2-shape .
the lines are the iv curves ( showing hysteresis ) .
the dashed lines indicate the bias , voltage for the @xmath0-shape resistor , current for the @xmath2-shape case .
the short - dashed lines show the switching from a branch to another when the bias is reversed /current bias for the @xmath0-shape , voltage bias for the @xmath2-shape .
finally , @xmath4 and @xmath5 represent the dip of the negative differential resistor , see sections [ sshape ] and [ figure].,title="fig:",width=264 ] .
, height=226 ] fig .
[ fig1]a shows schematically both the ndr iv curve of a semiconductor gunn diode @xcite ( which is used as a microwave source ) as well as that of a josephson junction coupled to a cavity @xcite .
this type of iv curve is sometimes referred to as an @xmath0-shaped iv curve @xcite .
the analogy with gunn diode is purely hypothetical .
we speculate that the ndr iv characteristic , even if of very different nature , could be similar to josephson junction ndr at a phenomenological level .
this is only an observation , and we do not claim that the constitutive physics is the same . in order to measure the ndr part of the iv curve in fig .
[ fig1]a , we must have a constant voltage bias source . with a constant current bias source we will see hysteresis when increasing and decreasing bias current ( dashed line ) . for
the josephson junction the cavity may typically look as the hysteretic curve if the junction is current biased or as the continuous curve ( full line ) when voltage bias . in most cases
the experimental curves for josephson junctions are current biased although voltage bias can also occur @xcite .
figure [ fig1]b shows another type of ndr curve referred to as @xmath2-shaped . here a constant current bias source is needed to trace out the full curve . in reference
@xcite a ndr region qualitatively similar to that in fig .
[ fig1]b is observed together with radiation emission . since additionally in @xcite radiation emission
is also observed in a part of the iv curve that looks qualitatively similar to that in fig .
[ fig1]a , it is tempting to link the radiation emission with the ndr .
we note that a josephson junction is nonlinear up to very high frequencies ( thz ) , implying that it s negative resistance may have important high frequency applications : examples are ( i ) a high frequency parametric amplifier and ( ii ) a generator for high frequency electromagnetic radiation .
we note from fig .
[ fig1 ] that we have 4 different combinations of the bias and @xmath6 nonlinearity . in the general case
the bias is neither current nor voltage biased but has a load line that is neither vertical nor horizontal .
however , a pure voltage bias is not conceivable for a superconducting element , and therefore there is not a full symmetry between the two cases , as can be seen in the following .
for intrinsic josephson junctions the ndr qualitatively shown in fig .
[ fig1]b comes from the well known back bending of the iv curve near the energy gap due to heating .
there are other shapes than @xmath2-shaped and @xmath0-shaped , an example from semiconductors is the @xmath7-shape a diagram showing the model we are discussing is depicted in fig .
[ fig2 ] . here
the left part of the figure shows the two different bias circuits , the middle part ( b ) shows the josephson junction model with the nonlinear resistor and the capacitor .
we do not hypothesize on the physical origin responsible for the ndr .
instead , a phenomenological iv characteristic of the effective shunt encompasses the behavior of the interaction with an external load .
shortly , we adopt the following logic : the observed iv of loaded stacks exhibits a ndr , therefore we introduce a nonlinear effective shunt resistor . for the bias circuit we distinguish between ( a ) current bias and ( c ) voltage bias . in the case of current bias
the shunt resistance @xmath8 is large and for the voltage bias case the bias resistor @xmath9 is small .
we underline that @xmath10 leads to an unphysical result for a superconducting element as the josephson junction .
in fact a dc voltage produces an ac current @xmath11 through the josephson junction element , while the capacitor remains inactive and the resistor current is constant .
the actual bias at high frequency is more complicated than elements ( a ) and ( c ) of fig .
[ fig2 ] @xcite , and we use such very simple schemes for sake of simplicity .
we notice that in experiments a partial voltage bias has been employed ( e.g. , @xcite ) , and a voltage bias is sometimes used in modeling ( e.g.,@xcite ) ; the interface between the voltage source and the mesa at thz frequencies leads to a very difficult full wave electromagnetic problem . in the following
we shall set up a mathematical model for the josephson junction circuits depicted in fig .
the voltage across the josephson junction is denoted @xmath12=@xmath13 and it depends on time @xmath14 .
the cooper pair phase difference across the josephson barrier is @xmath15=@xmath16 and the josephson voltage relation reads @xmath12=@xmath17 d\phi / dt$ ] .
the electron charge is @xmath18 and @xmath19 denotes planck s constant divided by @xmath20 . .
we model the @xmath0-type ndr characterized by a nonlinear conductance @xmath21 shaped as a gaussian function @xmath22 \right\ } \ ; .\ ] ] here @xmath23 is the linear quasi particle conductance ( related to the quasi particle resistance @xmath24 ) and @xmath25 describes the nonlinear part of the conductance .
the parameter @xmath4 position the gaussian function along the @xmath12-axis and @xmath26 is a measure of the gaussian width .
the gaussian model is phenomenological it is only a way to create the negative differential resistance .
in fact , we have also tried with other functional shapes ( e.g. polynomial ) , obtaining the same qualitative result : an increase of the power associated with the ndr branch , as in @xcite . a detailed comparison that could give a quantitative evaluation of the parameters @xmath25 and @xmath26 is out of the scope of this paper , which aims to a phenomenological description . the quasi particle tunnel current @xmath27 is then given by @xmath28 in fig .
[ fig2 ] @xmath29 denotes the bias current supplied by the current source in the circuit ( a ) or voltage source in the circuit ( d ) , @xmath30 is the displacement current through the capacitor and the josephson tunnel current is denoted @xmath31 = @xmath32 , where @xmath33 is the josephson critical current .
the current balance becomes @xmath34 . inserting the josephson voltage relation and eq .
( [ iqp ] ) into the current balance equation gives @xmath35 the above eq .
( [ rsj1 ] ) can be transformed into normalized units using the scaling @xmath36 , with @xmath37 , thereby we obtain @xmath38 here the nonlinear conductance @xmath21 has been normalized and leads to the nonlinear dissipation term with @xmath39 in @xmath40 \right\ } \ ; .\ ] ] in the above expression @xmath41 , @xmath42 , and @xmath43 is the normalized voltage .
the equations reduce to the usual rsj model for @xmath44 .
model eq.([rsjnorm ] ) is a phenomenological equation for the ndr observed in the full model @xcite : @xmath45 where @xmath46 is the tridiagonal matrix that describes the coupling among the superconducting layers @xcite .
the matrix reads @xmath47 in the diagonal , and the coupling constant @xmath2 otherwise . in the following
we will adopt the phenomenological model ( [ rsjnorm ] ) that qualitatively describes the ndr obtained by the full model ( [ stacka],[stackb ] ) ( see ref .
@xcite ) .
-type resistor .
crosses ( red symbols ) indicate increasing and decreasing bias current .
the available power is estimated through the maximum voltage oscillations .
parameters of the simulations are @xmath48 , @xmath49 , @xmath50 , @xmath51 , @xmath52.,width=226 ] -type resistor . crosses ( red symbols ) indicate increasing and decreasing bias current .
the available power is estimated through the maximum voltage oscillations .
parameters of the simulations are @xmath48 , @xmath49 , @xmath50 , @xmath51 , @xmath52.,width=226 ]
inserting a @xmath0-shaped resistor in @xmath53 , see eq.([gnorm ] ) , leads to current voltage curves of the form depicted in fig .
[ fig_ibias_nshape ] . the model for the @xmath0-shaped resistor
can also be used in the case of voltage bias , where the normalized current reads ( see fig .
[ fig2 ] ) : @xmath54 where @xmath55 is the normalized resistance @xmath9 of fig .
[ fig2]c . the expression in ( [ gbias ] )
can be inserted into the system equation ( [ rsjnorm ] ) , which reduces to the current bias when @xmath56 and @xmath57 . in this setup
we consider open circuit boundary conditions .
more realistic boundary conditions as an rlc circuit will be discussed elsewhere @xcite . in fig .
[ fig_ibias_nshape ] the available power is estimated through the oscillations of the josephson voltage .
we underline that in this paper we refer to _ available power _ , or shortly _
power _ ( also in the figure labels ) , as the maximum power that could be extracted supposing a perfect matching with the external device .
the optimization of this important effect is relevant for the application , but in this paper we focus on the existence of a general effect due to ndr .
the power estimate is done recording the jj waveforms and extracting the difference between the maximum and the minimum , that it is proportional to the mean square amplitude in the sinusoidal regime .
the features of the power estimate are generic and consistent with a full fourier analysis of the waveform .
the nonsinusoidal response of a single jj at low bias is enhanced and results in a high harmonic content at low bias , close to the return point on the iv .
the voltage oscillations seems to go towards infinity at low dc voltage .
as it is well known @xcite , the actual power is peaked at some ( low ) frequency and goes towards zero at high frequencies ( due to capacitance ) and low frequencies ( due to inductance ) .
it is also known that there is a phase change in the ac - voltage ( respect to the ac current ) as we decrease bias current , such that @xmath58 will not increase . moreover , in our calculations we only find the rising branch that is stable , and that abruptly switches to zero voltage at a finite level of amplitude ( corresponding to a finite current before the switch to the josephson supercurrent ) . applicationwise the most interesting region is above the peak when the voltage , and hence the frequency of the emitted radiation , span the thz region . in summary , from figs .
[ fig_ibias_nshape ] , [ fig_vbias_nshape ] it is evident that : ( i ) the nonlinear mccumber branch causes a hysteretic behavior .
( ii ) the output power , i.e. the amplitude of the josephson voltage oscillations , is nonlinear in the region of mccumber nonlinearity .
( iii ) the negative branch of the nonlinearity ( region i in fig .
[ fig_vbias_nshape ] , corresponds to a power higher than the power obtained from a linear mccumber resistor .
this is only evident in the case of voltage bias , for the nonlinear part is unstable with a current bias .
( iv ) the reverse is true for the positive branch of the nonlinearity ( region ii in fig .
[ fig_vbias_nshape ] ) , where the available power is lower than the power obtained from a linear mccumber resistor .
an @xmath2-type ndr may also be found in a josephson junction .
the most obvious example is the back bending of the iv curve in an intrinsic josephson junction at the energy gap @xcite .
this case is mathematically more complex than the @xmath0-type . for the @xmath0-type case we noticed that to have a continuous iv curve with ndr we should have voltage bias ( fig . [
most experiments on josephson junctions are current biased and give rise to hysteresis ( see fig .
[ fig_ibias_nshape ] ) . for @xmath2-type nonlinearity
the situation is the inverse .
the continuous iv curve is obtained with current bias whereas with voltage bias switching will result ( fig [ fig1]b ) . if we consider a current biased josephson junction circuit with an @xmath2-type nonlinearity ( see fig .
[ fig1]b ) we obtain for a nonlinear resistor : @xmath59 \right\ } \\
i_r & = & \frac{\hbar}{2e}g(i_r)\frac{d\phi}{d t}. \label{irs}\end{aligned}\ ] ] ) for an @xmath2-type ndr model.,width=264 ] that reduces to the linear case for @xmath60 . as usual with the current balance one gets the set of equations : @xmath61 inserting the expressions for the current through the capacitor ( @xmath30 ) and the josephson element ( @xmath31 ) one retrieves the second order set of differential equations : @xmath62 that can be cast in normalized units ( with the standard definitions of josephson angular velocity @xmath63 and dissipation @xmath64 ) in a system of two equations : @xmath65 here the nonlinear conductance @xmath66 has been normalized and gives rise to nonlinear dissipation @xmath67 : @xmath68 \right\ } \ ; , \ ] ] where @xmath69 .
the equations can be finally cast in normal form with the variables @xmath15 and @xmath70 : @xmath71 \nonumber \\ \frac{d\phi}{d\tau } & = & \frac{\gamma_r}{\alpha\left ( \gamma_r \right ) } \ ; .\end{aligned}\ ] ] the equations reduce to the usual rsj model for @xmath44 : -type resistor and voltage biased . crosses ( red symbols ) indicate increasing bias current and circles ( green symbols ) depict the case of decreasing bias current .
the available power is estimated through the maximum voltage oscillations .
parameter of the simulations are : @xmath72 , @xmath73 , @xmath74 , @xmath50 , @xmath52 .
, width=264 ] -type resistor and voltage biased .
crosses ( red symbols ) indicate increasing bias current and circles ( green symbols ) depict the case of decreasing bias current .
the available power is estimated through the maximum voltage oscillations .
parameter of the simulations are : @xmath72 , @xmath73 , @xmath74 , @xmath50 , @xmath52 . , width=264 ]
the dissipation is employed to write the system of differential equations for the variables @xmath15 and @xmath70 that is true for any functional dependence of the nonlinear dissipation upon the resistor current @xmath70 @xmath75 \nonumber \\
\frac{d\phi}{d\tau } & = & \frac{\gamma_r}{\alpha\left ( \gamma_r \right ) } \ ; .\end{aligned}\ ] ] in general , for any @xmath2-shaped conductivity we can define @xmath76 as the ratio between the bias and the dissipation @xmath77 the system of equations ( [ rsj_s_general ] ) can be written : @xmath78 \nonumber \\
\frac{d\phi}{d\tau } & = & \frac{\gamma_r}{\alpha \left ( \gamma_r \right ) } \ ; .\end{aligned}\ ] ] the coefficient of the right hand side is the nonlinear differential conductance . unfortunately for an @xmath2-shaped iv curve this is a term that vanishes in two points , see fig .
[ lip ] . for the model ( [ sigma ] ) , of the analytical expression of @xmath79 ( see eq.([sigma ] ) ) , reads : @xmath80 the system becomes singular ( non lipschitzian ) for two values of the current , where the differential conductance vanishes .
also , for negative values the system can not be integrated @xmath81 therefore , for the @xmath2-shaped case , one branch of the iv ( and the corresponding power ) can not be reproduced , see fig .
[ fig_ibias_sshape ] . a linear perturbation analysis of eq.([rsj_s_sigma ] ) around the singular points ( [ sigma_non_lip ] ) , can be accomplished setting @xmath82 : @xmath83 eqs .
( [ cond_int ] ) predict that for small @xmath84 ( in the proximity of the singular point ) the differential equations are not well defined . in the figures for the @xmath2-shaped resistor in fact squares
denote the point implicitly defined by such condition , that well agree with the point where the differential equation can not be numerically integrated . inserting an @xmath2-shaped resistor in @xmath85
one gets in the simulations a typical result in fig .
[ fig_ibias_sshape ] . also for the @xmath2-shaped resistor we can assume a voltage bias , see fig .
[ fig_vbias_sshape ] .
the normalized current is the same as in eq .
( [ gbias ] ) , that inserted into the system equation reads : @xmath86 \cdot \nonumber \\ & & \left [ \frac{\alpha_0}{r_b}\left ( \frac{v}{v_n}-\frac{d\phi}{d\tau}\right ) -\gamma_r - \sin(\phi ) \right ] \nonumber \\
\frac{d\phi}{d\tau } & = & \frac{\gamma_r}{\alpha\left ( \gamma_r \right ) } \ ; .\end{aligned}\ ] ] the system reduces to the current bias when @xmath87 and @xmath88 .
we note that there are no stable bias points between @xmath89 and @xmath90 in figs .
[ fig_ibias_sshape ] and [ fig_vbias_sshape ] . increasing the bias current from zero in fig .
[ fig_ibias_sshape ] we first have a supercurrent until the current is @xmath47 where we have a switching to the lower branch of the iv curve .
increasing the bias current further we reach point @xmath90 . still increasing the bias current from @xmath90 we continue from point @xmath89 .
we note that the power ( @xmath58 ) dissipated in @xmath89 and @xmath90 are the same so the transition between @xmath89 and @xmath90 is rather unusual .
there are no stable bias points between @xmath89 and @xmath90 and it is not a usual switching along a vertical or horizontal line .
figures [ fig_ibias_sshape],[fig_vbias_sshape ] include josephson radiation in terms of voltage amplitude . in both figs .
[ fig_ibias_sshape ] and [ fig_vbias_sshape ] we also have unusual radiation patterns . decreasing the bias current from above in fig .
[ fig_ibias_sshape ] we see the following : we observe radiation corresponding to the standard mccumber curve , until we reach @xmath89 . from here
we continue down from @xmath90 and the radiation follows a new curve showing a peak deviating from the standard mccumber curve in region @xmath91 . increasing the bias current from zero we have the switching at @xmath92 , but
otherwise the radiation is the same as we had for decreasing bias current .
thus even if we have two branches of the radiation curves we do not have hysteresis in the usual sense . figure [ fig_vbias_sshape ] with constant voltage bias shows essentially the same features .
@xmath0-type and @xmath2-type nonlinearities have been investigated in the mccumber curve for point josephson junctions .
a fictitious ndr , gaussian shaped , perturbation of the standard mccumber dissipation is introduced to take into account the complicated electrodynamic interaction between stacks of jj and the load .
this allows us to obtian @xmath0-type and @xmath2-type nonlinearities that posses differential resistance .
we found that the current - voltage characteristics of the full rsj josephson junction model is very much influenced by a nonlinearly shaped mccumber curve leading to hysteresis and enhanced emitted radiation power in the region with negative differential resistance .
current - voltage curves were obtained for both current bias and voltage bias .
more realistic high frequency schemes of the voltage bias and of the lumped elements could be attempted in further investigations .
our model governs a simple point josephson junction , but the results can be applied to phase locked stacks of josephson junctions , e.g. coupled via a common rlc cavity circuit .
we thank cost program mp1201 nanoscale superconductivity : novel functionalities through optimized confinement of condensate and fields ( nanosc -cost ) for partial support .
gf work at the dtu was supported by the cost `` short term scientific mission '' cost - stsm - ecost - stsm - mp1201 - 291013 - 036429 .
vp thanks infn ( italy ) for partial support .
l. ozyuzer , a.e.koshelev , c.kurter , n. gopalsami , q. li , m. tachiki , k. kadowaki , t. yamamoto , h. minami , h. yamaguchi , t. tachiki , k. e. gray , w .- k .
kwok , and u. welp , `` emission of coherent thz radiation from superconductors , '' _ science _ , vol .
5854 , pp .
1291 - 1293 , 2007 .
h. b. wang , s. guenon , b. gross , j. yuan , z. g. jiang , y. y. zhong , m. grunzweig , a. iishi , p. h. wu , t. hatano , d. koelle , and r. kleiner , `` coherent terahertz emission of intrinsic josephson junction stacks in the hot spot regime , '' _ phys .
5 , pp . 057002 - 1 - 057002 - 4 , 2010 .
t. kashiwagi , m. tsujimoto , t. yamamot , h. minami , k. yamaki , k. delfanazari , k. deguchi , n. orita , t. koike , r. nakayama , t. kitamura , m. sawamura , s. hagino , k. ishida , k. ivanovic , h. asai , m. tachiki , r. a. klemm , and k. kadowaki , `` high temperature superconductor terahertz emitters : fundamental physics and its applications , '' _ japan
. j. appl .
010113 - 1 - 010113 - 14 , 2012 .
t. m. benseman , a. e. koshelev , k. e. gray , w .- k .
kwok , u. welp , k. kadowaki , m. tachiki , and t. yamamoto , `` tunable terahertz emission from bi2sr2cacu2o8+@xmath26 mesa devices '' _ phys .
b _ , vol . 84 , no .
064523 - 1 - 064523 - 6 , 2011 .
m. tsujimoto , k. yamaki , k. deguchi , t. yamamoto , t. kashiwagi , h. minami , m. tachiki , k. kadowaki , and r. a. klemm , `` geometrical resonance conditions for thz radiation from the intrinsic josephson junctions in bi2sr2cacu2o8+@xmath26 '' , _ phys .
037005 - 1 - 037005 - 4 .
2010 .
m. cirillo , n. grnbech - jensen , m. r. samuelsen , m. salerno , and g. verona rinati , `` fiske modes and eck steps in long josephson junctions : theory and experiments , '' _ phys .
b _ , vol . 58 , no .
12377 - 12384 , 1998 .
k. kadowaki , h. yamaguchia , k. kawamata , t. yamamoto , h. minami , i. kakeya , u. welp , l. ozyuzer , a. koshelev , c. kurter , k.e .
gray , w .- k .
kwok , `` direct observation of tetrahertz electromagnetic waves emitted from intrinsic josephson junctions in single crystalline bi2sr2cacu2o8+@xmath26 , '' _ physica c _ , vol .
7 - 10 , pp .
634 - 639 , 2008 . j. nagel , d. speer , t. gaber , a. sterck , r. eichhorn , p. reimann , k. ilin , m. siegel , d. koelle , and r. kleiner , `` observation of negative absolute resistance in a josephson junction , '' _ phys . rev .
217001 - 1 - 217001 - 4 , 2008 .
m. li , jie yuan , n. kinev , j. li , b. gross , s. gunon , a. ishii , k. hirata , t. hatano , d. koelle , r. kleiner , v. p. koshelets , h. wang , and p.
wu , `` linewidth dependence of coherent terahertz emission from bi2sr2cacu2o8+@xmath26 intrinsic josephson junction stacks in the hot - spot regime , '' _ phys . rev .
060505(r ) , 2012 .
b. gross , j. yuan , d. y. an , m. y. li , n. kinev , x. j. zhou , m. ji , y. huang , t. hatano , r. g. mints , v. p. koshelets , p. h. wu , h. b. wang , d. koelle , and r. kleiner , `` modeling the linewidth dependence of coherent terahertz emission from intrinsic josephson junction stacks in the hot - spot regime , '' _ phys . rev .
88 , pp . 014524 , 2013 .
pedersen , g. filatrella , v. pierro , and m.p .
srensen , `` negative differential resistance in josephson junctions coupled to a cavity , '' subm . to physica c , proceedings of _ viii int .
conf . on vortex matter in nanostrucutred superconductors _ ,
rhodes ( greece ) , 21 - 26 september , 2013 . | josephson junction systems with a negative differential resistance play an essential role for applications . as a well known example , long josephson junctions of the bscco type have been considered as a source of thz radiation in recent experiments .
numerical results for the dynamics of the fluxon system have demonstrated that a cavity induced negative differential resistance plays a crucial role for the emission of electromagnetic radiation .
we consider the case of a negative differential resistance region in the mccumber curve itself of a single junction and found that it has an effect on the emission of electromagnetic radiation .
two different shapes of negative differential resistance region are considered and we found it is essential to distinguish between current bias and voltage bias .
+ + _ keywords : millimeter wave devices , superconducting devices , nonlinear circuits , nonlinear optics , nonlinear oscillators _ |
over the last several decades , the standard model of weak interactions has been thoroughly tested by a vast amount of data for leptonic and semi- leptonic decays and reactions .
hadronic weak interactions are in general more difficult to study experimentally since they are usually obscured by the presence of the much larger strong interaction .
this requires employing processes in which the strong force can not participate due to overriding symmetry principles . in the case of the weak nucleon - nucleon interaction
it was realized more than 40 years ago@xcite that the current - current form of the weak interaction dictates the presence of a weak transition between nucleons which would lead to parity impurities in nuclear states which are of first order in the weak coupling .
using these parity nonconserving observables in many experiments on nuclear gamma and alpha transitions , polarized @xmath2 scattering , as well as the recent first measurement of the nuclear anapole moment@xcite , much has been learned about the weak nucleon - nucleon interaction@xcite .
the situation is very different for the flavor - changing baryon - baryon interaction .
soon after the discovery of hypernuclei it was recognized that the @xmath5 inside the nuclear medium decays not through its pauli - blocked mesonic decay channel , but predominantly through the @xmath11 = 1 nonmesonic transition @xmath12 , thus opening a door to the study of the weak strangeness - changing hyperon - nucleon force .
however , experimental progress in this field of weak hypernuclear decays has been slow until recently due to the difficult multi - coincidence , low count - rate nature of these measurements . in recent years
the situation has improved significantly due to a series of new experiments at bnl and kek . on the theoretical side it became clear that in order to unambiguously extract the weak @xmath12 transition potential significant effort
must be spent to account for the nuclear structure effects as accurately as possible .
when this task was recently completed@xcite it became apparent that major discrepancies between theory and experiment can not be due to the underlying nuclear structure but have to arise from the nature of the weak transition potential itself .
even with the nuclear structure input under control it has become desirable to measure the process @xmath0 directly since the hypernuclear decay can only probe the reaction at one well - defined kinematic setting . though considered impossible for many years since
this process is reduced in cross section by around 12 orders of magnitude from the standard elastic @xmath2 scattering , recent progress in experimental and accelerator technology may have brought measuring this process within reach @xcite .
it is therefore timely to provide predictions of various observables based on the transition potential used in the nonmesonic hypernuclear decay . in sec .
[ dcs ] of this paper , we present the expressions for the matrix elements and the cross section . sec .
[ wkp ] briefly describes the transition operator derived in ref.@xcite .
our results are discussed in sec .
[ res ] and summarized in sec .
the differential cross section per unit solid angle in the center - of - mass system for the reaction @xmath13 as depicted in fig .
[ fig : diag1 ] is given by the expression @xmath14 with @xmath15 the total available energy in the center - of - mass system and @xmath16 and @xmath17 the relative momenta of the particles in the initial and final states respectively . in a plane wave
born approximation ( pwba ) the weak transition matrix elements read @xmath18 where the overline stands for the antisymmetric combination of the two particle states @xmath19 and @xmath20 and @xmath21 is the nonrelativistic weak transition potential .
accounting for the locality of the weak potential , the direct term of the previous matrix elements can be written in the distorted wave born approximation ( dwba ) as @xmath22^ * { \chi^\dagger}^{t}_{m_t } v^{w}({\vec r}\ , ) \psi_{nn}^{(+)}({\vec p}_i,{\vec r } ) \chi^{t}_{m_t } \ ] ] where @xmath23 contains the radial , angular and isospin dependence of the weak transition potential and @xmath24 ( @xmath25 ) stands for the distorted @xmath26 ( @xmath2 ) wave function . in sec .
[ wkp ] it is shown how the weak potential can be decomposed as @xmath27 where the index i sums over mesons and @xmath28 over the different spin channels .
the radial part of the potential is denoted by @xmath29 , the piece containing the angular and spin dependence by @xmath30 and @xmath31 denotes the appropriate isospin operator for each meson . using the partial wave decomposition for the distorted waves , working in the coupled basis formalism , @xmath32 ( see appendix ) , and assuming @xmath33 parallel to the z - axis , the modulus squared of the weak matrix elements for the @xmath13 reaction finally reads @xmath34^j_{l_f ' s_f ' , l_f s_f } ( k_f , r ) \,v_{\alpha}^{(i)}(r ) \,[\psi^{(+)}_{nn}]^j_{l_i ' s_i , l_i s_i } ( k_i , r ) \nonumber \\ & \times & \left .
\int d\omega \,{{\cal j}^{\dagger}}^{j m}_{l_f ' s_f ' } ( { \hat r } ) \,v_{\alpha}^{(i)}({\hat r } ) \,{\cal j}^{j m}_{l_i ' s_i ' } ( { \hat r } ) \,\,\left(\frac{1-(-1)^{l_i'+s_i+t}}{\sqrt{2 } } \right ) \right|^2 .
\label{eq : mef}\end{aligned}\ ] ] the distorted radial wave functions in the above equation are generated from a t - matrix which is constructed using the nucleon - nucleon ( @xmath2 ) and hyperon - nucleon ( @xmath3 ) strong potentials .
we make use of the nijmegen 93 @xcite and bonn b @xcite @xmath2 potentials and the nijmegen soft - core @xcite and jlich @xcite @xmath3 potentials .
comparison between the results obtained using the different interaction models is made in sec .
in ref . @xcite a one - boson - exchange model is developed to describe the @xmath1 transition , where the pseudoscalar @xmath35,k and vector @xmath36 and k@xmath37 mesons mediate the interaction .
we use this model in order to describe the present inverse reaction . in this study
we have refrained from considering the transition @xmath38 , since it was found to be an order of magnitude smaller than that for the direct @xmath5 production @xcite .
the nonrelativistic reduction of the feynman amplitude , corresponding to the diagram depicted in fig .
[ fig : diag1 ] , leads to the nonrelativistic weak potential in momentum space , which for the exchange of pseudoscalar mesons takes the form @xmath39 where @xmath40 is the fermi coupling constant , @xmath41 is the momentum carried by the meson ( m ) directed towards the strong vertex , @xmath42 the strong coupling constant for the nnm vertex , @xmath43 the meson mass , @xmath44 the nucleon mass and @xmath45 the average between the nucleon and @xmath5 masses . the operators @xmath46 and @xmath47 contain , in addition to the weak coupling constants , the particular isospin structure corresponding to the exchanged meson . in the case of vector meson exchange
the weak potential takes the form @xmath48 with @xmath49 , @xmath50 the strong coupling constants and @xmath51 , @xmath52 and @xmath53 the weak coupling constants that also contain the appropriate isospin operator of the particular meson . performing a fourier transform of eqs .
( [ eq : pspot ] ) and ( [ eq : vpot ] ) , and using the relation @xmath54 ( _ 2 q)@xmath55 \cdot \hat{r } \right ) v_{pv}^{(i)}(r ) \hat{i}^{(i)}_{pv } \right ] \ , \label{eq : genpot } \end{eqnarray } where the index $ ] i@xmath56i=1 , , 6@xmath57,,@xmath58,,@xmath59^*@xmath60@xmath61c@xmath62ss@xmath63t@xmath64pv@xmath65n^i = 1 ( 0)@xmath66@xmath67@xmath68_1 @xmath69 and for isoscalar mesons ( @xmath70,@xmath71 ) this factor is just @xmath72 for all spin structure pieces of the potential .
in the case of isodoublet mesons ( k , k@xmath37 ) there are contributions proportional to @xmath72 and to @xmath73 that depend on the coupling constants and , therefore , on the spin structure piece of the potential denoted by @xmath28 .
for k - exchange we have @xmath74 and for k@xmath37-exchange @xmath75 the different pieces @xmath29 , with @xmath76 , are given by @xmath77 \equiv k^{(i)}_{ss } v_{ss } ( r,\mu_i ) \label{eq : sspot } \\
v^{(i)}_{t } ( r ) & = & k^{(i)}_{t } \ : \frac { 1}{3 } \ : \mu_i^2 \ : \frac { { \rm e}^{- \mu_i r } } { 4 \pi r } \ : \left ( 1 + \frac{3 } { { \mu_i } r } + \frac{3 } { ( { \mu_i r})^2 } \right ) \equiv k^{(i)}_{t } v_{t } ( r,\mu_i ) \label{eq : tpot } \\
v_{pv}^{(i)}(r ) & = & k^{(i)}_{pv } \ : \mu_i \ : \frac { { \rm e}^{- { \mu_i } r } } { 4 \pi r } \left ( 1 + \frac{1 } { \mu_i r } \right ) \equiv k^{(i)}_{pv } v_{pv}(r,\mu_i ) \ .
\label{eq : pvpot}\end{aligned}\ ] ] where @xmath78 denotes the mass of the different mesons .
the expressions for @xmath79 , which contain factors and coupling constants , are given in table [ tab : ctspot ] .
the explicit values of the coupling constants can be found in table [ wkp ] of ref .
@xcite .
a monopole form factor @xmath80 is used at each vertex , where the value of the cut - off , @xmath81 , depends on the meson .
these values are the same ones as those of the strong jlich @xmath3 interaction@xcite , since the nijmegen model distinguishes form factors only in terms of the transition channel .
the resulting expressions for the regularized potentials , which were given already in ref .
@xcite , will be repeated here in order to correct for some misprints .
the effect of form factors is included by making the following replacements in eqs .
( [ eq : cpot ] ) to ( [ eq : pvpot ] ) @xmath82
the total cross section for the @xmath13 reaction including only one - pion - exchange in the weak transition potential is shown in fig . [
fig : fig1 ] as a function of the proton lab momentum , starting at the @xmath83 threshold . as expected for a weak process
, the cross section is of the order of 10@xmath84 mb .
the dashed line corresponds to the free " calculation where form factors are omitted at the vertices and the incoming @xmath85 and outgoing @xmath86 states propagate as plane waves .
the thin solid line includes form factors and strong initial @xmath85 distorted waves obtained from the nijmegen 93 @xmath2 interaction @xcite .
the effect of the @xmath2 interaction is to reduce the cross section by a factor of about 2 .
the reduction is a result of the fact that the @xmath85 pair moves with large relative momentum ( from @xmath87 mev / c to @xmath88 mev / c ) in the range of proton lab momenta explored and feels the effect of the strong repulsive short range part of the @xmath2 interaction . as a consequence , the @xmath85 wave function
is much reduced at small distances where the weak transition potential contributes most to the cross section and hence the cross section becomes smaller .
the thick solid line shows the cross section when the @xmath83 distortions , obtained from the @xmath3 nijmegen soft core potential @xcite , are also included .
when compared to the thin line , one observes a substantial enhancement of the cross section in the low momentum region which turns into a moderate reduction at high lab momentum values .
this is due to the fact that , close to threshold , the @xmath83 pair moves with small relative momentum , feeling the attractive component of the @xmath3 potential . as a result
the wave function gets pushed in to smaller distances and the cross section is enhanced . as the proton lab momentum increases ,
so does the relative @xmath89 momentum and , eventually , the repulsive core of the @xmath3 interaction becomes responsible for reducing the wave function at short distances , giving rise to smaller cross sections .
the opening of the @xmath90 channel , coupled to the @xmath91 one through the strong interaction , shows up as a step in the cross section at a proton lab momentum around @xmath92 mev / c .
the effect of including the different mesons in the weak transition potential is shown in fig .
[ fig : fig2 ] . fig .
[ fig : fig2]a shows the cross sections calculated using the nijmegen potentials for getting the @xmath2 and @xmath26 distorted waves and fig .
[ fig : fig2]b shows those using the bonn b @xmath2 and jlich a @xmath26 potentials . in fig .
[ fig : fig2]a we see that adding the @xmath93 meson to one - pion - exchange considerably reduces the cross section in the kinematic region studied here .
including the remaining mesons does not change the cross section much , which ends up being about a factor 2 smaller than that for one - pion only .
this effect appears to be surprisingly different from the moderate reduction of 15% found for the decay rate of hypernuclei when the effect of adding heavier mesons to the one - pion - exchange mechanism was studied @xcite .
however , we point out that the @xmath26 correlations in the weak hypernuclear decay case were obtained from a g - matrix calculation which takes into account the pauli principle acting on the intermediate nucleon .
this blocks the low relative momentum transitions between the initial @xmath26 and intermediate @xmath3 states and , therefore , the effect of the attractive part of the interaction on the correlated @xmath26 wave function is much reduced compared to those of the distorted @xmath89 wave function in free space .
thus , the @xmath26 wave function in the hypernucleus is strongly suppressed at small distances , making the contribution of the heavier mesons less important . as can be seen in fig .
[ fig : fig2]b , the use of the bonn b @xmath2 and jlich a @xmath26 potentials give rise to different effects in the cross section as compared to those calculated using the nijmegen ones .
although not shown in the figures , we have checked that , in the absence of @xmath89 distortions , the cross section calculated with the distorted @xmath94 waves , obtained from either the nijmegen 93 or the bonn b potentials , yield nearly identical results .
therefore , the differences between the results in figs .
[ fig : fig2]a and [ fig : fig2]b come almost completely from the different models used to distort the @xmath83 final state . already at the level of only pion - exchange ( dashed line in fig .
[ fig : fig2]b ) , the jlich a model shows a clear enhancement close to the @xmath95 threshold . adding the @xmath93 meson furthermore yields a strong reduction of the cross section . on the other hand ,
the addition of the remaining mesons produces a substantial enhancement which gives rise to a final cross section not very different from the pion - only result .
we note here that the pion - only cross section shown in fig .
[ fig : fig2 ] is close to the results obtained in ref .
@xcite with one - pion - exchange , but we do not find the tremendous enhancement of a factor of 3 in the cross section when the @xmath93 meson is included .
this is most likely due to the different models for the weak @xmath96 vertices used in the @xmath93-exchange mechanism .
their weak coupling constants are larger than the ones used here by a factor of more than 2 for the parity conserving ones ( @xmath97 and @xmath98 ) and a factor 3.5 for the parity violating one ( @xmath99 ) .
let us now explore the origin of the peak in the cross section around the @xmath6 threshold .
nuclear matter calculations already showed some time ago @xcite that , while the @xmath5 binding energy in nuclear matter turned out to be about @xmath100 to @xmath101 mev for the nijmegen and jlich interactions , the distribution of this strength between the various partial waves was very different .
the most important contribution for the jlich interactions came from the coupled @xmath102 channel , the @xmath103 contribution being negligible .
however , the nijmegen interactions obtained almost equal contributions from both the @xmath104 and @xmath105 channels . in fig .
[ fig : fig3 ] we analyze the contribution to the @xmath106 cross section from @xmath102 partial waves obtained from the nijmegen ( fig . [
fig : fig3]a ) and jlich a ( fig .
[ fig : fig3]b ) @xmath91 interactions . the thick solid lines ( `` _
full _ '' ) are the calculated cross sections with one - pion - exchange for the weak transition @xmath107 and including distorted waves for the @xmath85 and @xmath86 systems .
the short dashed lines represent cross sections obtained by removing the @xmath108 component of the distorted @xmath83 wave function and the long - dashed lines those obtained when the @xmath109 component is removed .
the thin solid line shows the small cross sections obtained when all the @xmath102 coupled channel is omitted , thus reflecting the fact that these are the most important components for the @xmath110 reaction , as already noted in ref .
the stronger @xmath102 channel of the jlich interaction , relative to the nijmegen one , is the reason for the much more enhanced cross section at the @xmath111 threshold . suppressing the @xmath103 partial waves
barely affects the cross section and can not be seen in figs .
[ fig : fig3]a and [ fig : fig3]b . in spite of the fact that the nijmegen @xmath3 interaction has a sizable @xmath103 component , its contribution to the cross section
gets eliminated due to the weak one - pion - exchange transition potential which has negligible strength for the @xmath112 transition . in fig .
[ fig : fig4 ] we show the differential cross sections for the proton lab momentum of @xmath113 mev / c which corresponds to the position of the peak in fig .
[ fig : fig2 ] , where the strong @xmath2 and @xmath26 correlations have been generated with the bonn b and jlich a interactions , respectively .
the dashed line is the result for the pion - only calculation , the thin solid line shows the effect of including the @xmath93 meson and the thick solid line incorporates the full set of six mesons .
the effect of adding the rho to the pion is a reduction of a factor slightly more than 2 while adding all mesons brings the results closer to the one - pion ones , as observed earlier for the total cross section . a measure of the amount of parity violation in the weak @xmath5 production is given by the asymmetry @xmath114 defined as @xmath115 where @xmath116 ( @xmath117 ) is the cross section for positive ( negative ) helicity of the incoming proton . in fig .
[ fig : fig5 ] we consider the one - pion - exchange mechanism in the weak @xmath13 transition and illustrate the effect of the strong distortions on the asymmtery a using the nijmegen models for the @xmath2 and @xmath26 interactions .
the dashed line shows the asymmetry when neither the form factors at the vertices nor the strong distortions are included .
the inclusion of the @xmath2 interactions , as shown by the thin solid line , decrease the value of the asymmetry up to a beam momentum of 1130 mev / c beyond which it increases compared to the dashed line .
the addition of the @xmath26 final state interaction gives rise to the particular cusp structure due to the opening of the @xmath90 channel around 1140 mev / c .
the effect of the different meson exchanges in the weak transition potential is shown in fig .
[ fig : fig6 ] where the @xmath2 and @xmath26 distorted waves have been generated with the bonn b and jlich a models , respectively .
we observe that while the @xmath93 meson increases the asymmetry with respect to the pion - only result , the inclusion of all the six mesons produces an asymmetry which is considerably smaller .
similar qualitative features are found when the nijmegen potential models are used in constructing the distorted @xmath2 and @xmath26 waves . finally , we present in figs . [
fig : fig7]a and [ fig : fig7]b the @xmath13 cross sections and asymmetries , respectively , obtained with the full meson - exchange weak transition potential for three different strong interaction models .
the dashed lines use the @xmath2 nijmegen 93 and @xmath3 nijmegen soft core models , the thin solid lines use the @xmath2 bonn b and @xmath3 jlich b models and the thick solid lines use the @xmath2 bonn b and the @xmath3 jlich a models . as mentioned before , the differences between the curves are mainly due to the differences in the @xmath3 potentials employed .
it is clear that these strong @xmath3 potentials are not sufficiently constrained by the small amount of total cross section data on @xmath3 scattering .
hence the different @xmath3 models which produce the total cross sections for @xmath3 scattering equally well , give rise to very different predictions when applied to other reactions that are sensitive to the @xmath3 interaction , like hypernuclear structure calculations , studies of nuclear matter with strangeness , or the weak transition @xmath13 studied here .
more data on @xmath3 scattering , especially on differential cross sections and polarization observables , is highly desirable in order to constraint the @xmath3 interactions sufficiently well so that the @xmath106 reaction can be used to learn about the weak four fermion interaction .
in particular , the two - step @xmath38 transition , not considered in the present work , could be studied after having a better knowledge of the @xmath3 interaction .
we have studied the weak strangeness production reaction @xmath118 in a distorted wave born approximation formalism using the one - meson - exchange model for the weak transition consisting of six mesons viz . the @xmath119k and k@xmath37 .
the distorted wave functions are written in terms of partial wave expansions and are generated using the different available potentials for nucleon - nucleon and hyperon - nucleon interactions .
the total cross sections are sensitive to the model ingredients of the weak transition operator . including the @xmath93 meson decreases the pion - only cross sections by a factor of 2 or more .
the effect of the remaining mesons depends on the strong potentials employed to distort the @xmath83 states , giving rise to cross sections that can be either very close or a factor 2 smaller than the pion - only results .
the kinematical region explored by the free @xmath120 reaction is much larger than that by the inverse reaction , the nonmesonic decay @xmath121 taking place inside hypernuclei .
the heavier mesons contribute very differently and are more important in the free reaction compared to the nonmesonic decay due to the different behavior of the @xmath26 wave function inside a hypernucleus compared to that in free space . the total cross section computed with the nijmegen @xmath26 wave functions
show a step - like behavior around 1140 mev / c beam momentum , where the strong @xmath122 transition opens up .
the jlich @xmath26 results show a dramatic peak in this region .
however , the peak in our results is not as pronounced as the one found in earlier work@xcite .
the major contribution to these cross sections comes from the @xmath26 partial waves in the @xmath102 coupled channels .
we find the @xmath120 reaction to be very sensitive to the type of model used for the strong hyperon - nucleon interaction .
hence , more data on @xmath3 scattering for observables other than the total cross sections are needed to constrain the @xmath3 interaction models and use the @xmath123 reaction to extract the weak four fermion interaction .
the cross sections for the @xmath120 reaction are of the order of 10@xmath84 mb and are at the borderline of feasibility for the existing experimental facilities .
this work is partially supported by the dgicyt contract no .
pb95 - 1249 ( spain ) , the generalitat de catalunya grant no .
grq94 - 1022 , the us - doe grant no .
de - fg02 - 95-er40907 and the nato grant no .
one of the authors ( n.g.k . ) would like to thank the warm hospitality of the nuclear physics group at the university of barcelona .
the lippmann - schwinger equation allows us to obtain the scattered wave function for a pair of particles moving under the influence of the strong interaction .
the scattered states will then be given by for the incoming @xmath2 states and by @xmath125 for the outgoing @xmath26 states .
the previous equations are written in terms of the t - matrix which is obtained from @xmath126 and fulfills @xmath127 @xmath128 and @xmath129 being the corresponding unperturbed states for the @xmath2 and @xmath26 systems , respectively .
note that eq .
( [ eq : tmat3 ] ) should have been written as @xmath130 however , since we disregard the @xmath131 transition , we do nt need to consider the @xmath132 component of the wave function , @xmath133 . simplifying the notation by writing @xmath134 as simply @xmath26 eq .
( [ eq : tmat32 ] ) finally reduces to eq .
( [ eq : tmat3 ] ) .
@xmath136^ * \,{\chi^\dagger}^{t}_{m_t } & = & \sqrt{\frac{2}{\pi}}\ , \sum_{j m } \,\sum_{l_f ' s_f ' } \,\sum_{l_f m_{l_f } } ( -i)^{l_f ' } \ , [ \psi^{(-)}_{\lambda n}]^{*\,j}_{l_f ' s_f ' , l_f s_f } ( p_f , r ) \nonumber \\ & \times&\langle l_f m_{l_f } s_f m_{s_f } | j m \rangle \,y_{l_f m_{l_f } } ( { \hat p}_f)\ , { { \cal j}^{\dagger}}^{j m}_{l_f ' s_f ' } ( { \hat r } ) \label{eq : wfln2}\end{aligned}\ ] ] and @xmath137^j_{l_i ' s_i , l_i s_i } ( p_i , r ) \nonumber \\ & & \langle l_i m_{l_i } s_i m_{s_i } | j m \rangle \,y_{l_i m_{l_i}}^ { * } ( { \hat p}_i ) \ , { \cal j}^{j m}_{l_i ' s_i } ( { \hat r } ) \label{eq : wfnn2}\end{aligned}\ ] ] where the quantities @xmath138 are the generalized spherical harmonics . the radial wave functions in the above equations are generated numerically using the t - matrix constructed from the appropriate potentials and are given as follows @xmath139^{*\,j}_{l_f ' s_f ' , l_f s_f } ( p_f , r ) & = & j_{l_f}(p_f ) \delta _ { l_fl_f ' } \delta _ { s_fs_f ' } \nonumber \\ & + & \left .
\int \frac { p'^2 dp ' \langle p_f(l_f s_f ) j | t_{\lambda n \to \lambda n } | p'(l_f's_f ' ) j \rangle}{e_n(p_f)+e_{\lambda}(p_f)-e_n(p')-e_{\lambda}(p')+i\eta } \ , j_{l_f'}(p'r ) \right.\end{aligned}\ ] ] @xmath140^{j}_{l_i ' s_i , l_i s_i } ( p_i , r ) & = & j_{l_f}(p_f ) \delta _ { l_il_i ' } \nonumber \\ & + & \left .
\int \frac { p'^2 dp ' \langle p'(l_i ' s_i ) j |
t_{nn \to nn } | p_i(l_is_i ) j \rangle}{2e_n(p_i)-2e_n(p')+i\eta } \ , j_{l_i'}(p'r ) \right
. .\end{aligned}\ ] ] r. feynman and m. gell - mann , phys . rev . *
109 * , 193 ( 1958 ) .
wood et al , science * 275 * , 1759 ( 1997 ) .
e. adelberger and w. haxton , ann .
nucl . part .
sci . * 35 * , 501 ( 1985 ) .
a. parreo , a. ramos , and c. bennhold , phys . rev .
c * 56 * , 339 ( 1997 ) .
t. kishimoto , _ weak and electromagnetic interactions in nuclei _ , eds . h. ejiri , t. kishimoto and t. sato , world scientific ( 1995 ) 514 .
stoks , r.a.m .
klomp , c.p.f .
terheggen and j.j .
de swart , phys .
c * 49 * , 2950 ( 1994 ) .
, eds . k. langanke , j.a . maruhn and s.e .
koonin , springer - verlag ( 1993 ) 1 - 29 .
maessen , th .
a. rijken and j.j .
de swart , phys .
c * 40 * , 2226 ( 1989 ) .
b. holzenkamp , k. holinde , and j. speth , nucl .
a500 * , 485 ( 1989 ) .
j. haidenbauer , k. holinde , k. kilian , t. sefzick and a.w .
thomas , phys .
c * 56 * , 3496 ( 1995 ) .
y.yamamoto , r. reuber , h. himeno , s. nagata and t. motoba , czec .
jour . phys . * 42 * , 1249 ( 1992 ) .
y.yamamoto , t. motoba , h. himeno , k. ikeda and s. nagata , prog .
. phys . * suppl .
117 * , 361 ( 1994 ) .
m. hjorth - jensen , a. polls , a. ramos and h. mther , nucl .
* a605 * , 458 ( 1996 ) . | the weak production of lambdas in nucleon - nucleon scattering is studied in a meson - exchange framework . the weak transition operator for the @xmath0 reaction is identical to a previously developed weak strangeness - changing transition potential @xmath1 that describes the nonmesonic decay of hypernuclei .
the initial @xmath2 and final @xmath3 state interaction has been included by using realistic baryon - baryon forces that describe the available elastic scattering data .
the total and differential cross sections as well as the parity - violating asymmetry are studied for the reaction pn @xmath4 p@xmath5 .
these observables are found to be sensitive to the opening of the @xmath6 production channel , the choice of the strong interaction potential and the structure of the weak transition potential .
# 1#2@xmath7#2 # 1eq .
( [ # 1 ] ) 12@xmath8c 11@xmath9b 5@xmath10he # 1#1| # 1| # 1 |
luminous blue variables are hot massive stars which undergo sporadic eruptions on timescales of years and decades @xcite .
the prototype is s doradus , as well as hubble - sandage variables in m31 and m33 @xcite , which shows eruptions of 1 - 2 magnitude level in a time - span of several decades .
other examples are @xmath1 carina and p cygni , which show giant eruptions ( @xmath2 2 mag ) at a frequency of several centuries .
@xcite is the first to coin the name luminous blue variables for this type of stars , and separates them from other type of bright blue stars like wolf - rayet stars .
lbvs play an important role at the very late stage of massive star evolution .
they are considered as a transition phase where o stars evolve toward wolf - rayet stars .
lbvs were originally regarded only as supernova impostors because they often show giant eruptions mimicking the explosion of supernovae , but the central star remains after the ejecta have been expelled .
however , a link between lbvs and supernova progenitors was suggested by when interpreting the radio lightcurves of supernovae . the radio emission seen after the supernova explosion is induced by the interaction between supernova ejecta and the progenitor s circumstellar medium , thus radio
lightcurves bear information on the mass - loss history of the progenitor . suggested that radio lightcurves of sne indicate high mass - loss histories of the progenitors which matches well with lbvs .
pre - eruption images of several sne also suggest lbvs as their progenitor .
for example , the progenitor of sn 1987a was recognized as a blue supergiant @xcite and @xcite suggested that it could be classified as a low - luminosity lbv .
@xcite identified the progenitor of sn 2005gl using _ hst _ and attributed it to be a lbv . recently
a previously known lbv - sn 2009ip - has undergone its third eruption and has been linked to a true supernova @xcite .
the nature of the recent eruption of sn 2009ip is under debate ; subsequent follow - up has been carried out to verify or reject it as a core - collapse sn @xcite .
yet there are only a few known lbvs , either in our galaxy or in m31 and m33 .
thus , increasing the number of known lbvs is essential toward understanding their nature and evolution .
in addition to the pioneering decades - long photometric monitoring campaign conducted by hubble and sandage @xcite , there are several methods to uncover lbvs .
for example , lbvs are strong uv and h - alpha emitters ( see * ? ? ? * and reference therein ) and can be revealed e.g. with observations of the _ galex _ satellite or h - alpha surveys .
@xcite conducted a h - alpha survey of m31 and m33 and followed - up a selected sample of strong h - alpha emitters spectroscopically . by comparing the spectra of their h - alpha emitter sample with known lbvs , they were able to identify candidate lbvs , which saved a substantial amount of time required to uncover lbvs photometrically .
because they have uncovered more than 2,500 h - alpha emitting stellar objects , they can only follow - up dozens of them , yet there are much more to be explored .
@xcite are currently exploring other h - alpha emitting sources in this list , in combination with infrared photometry including 2mass , _ spitzer _ and _ wise _ to search for luminous and variable stars .
since lbvs undergo several eruptions and exhibit high mass - loss rates , they accumulate vast amounts of material in their circumstellar environment which could be detectable in the infrared ( e.g. * ? ? ?
@xcite have made use of _ spitzer _ irac photometry and searched for @xmath1 carina analogs in nearby galaxies including m33 ( but not m31 ) .
they estimate that 6@xmath36 of their candidates are true @xmath1 carina - like sources . here
we outlined a novel approach utilizing mid - term photometric variation , as well as optical and infrared color to search for lbvs using the lggs optical and 2mass infrared photometry , with the combination of the photometric variability from pandromeda monitoring campaign .
our paper is organized as follows : in 2 we describe the optical and infrared data we use . in
3 we outline our method .
a discussion of our candidates is presented in 4 , followed by an outlook in 5 .
the time - series photometric data employed to search for variability are from the pandromeda project .
pandromeda monitored the andromeda galaxy with the 1.8 m ps1 telescope with a @xmath4 7 deg@xmath5 field - of - view ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for a detailed description of the ps1 system , optical design , and the imager ) .
observations have been taken in the @xmath0 and @xmath6 filters on daily basis during july to december between 2010 and 2012 in order to search for microlensing events and variables .
the distribution of the observations in the @xmath0-filter is shown in fig .
[ fig.cadence ] .
several exposures in @xmath7 , @xmath8 , and @xmath9 are also taken as complementary information for studies of the stellar content of m31 .
the data reduction is based on the mdia tool @xcite and is explained in @xcite in detail .
we outline our data reduction steps as follows .
the raw data are detrended by the image processing pipeline ( ipp , * ? ? ?
* ) and warped to a sky - based image plane ( so - called skycells ) .
the images at the skycell stage are further analyzed by our sophisticated imaging subtraction pipeline _
mupipe _ based on the idea of image differencing analysis advocated by @xcite .
this includes the creation of deep reference images from best seeing data , stacking of observations within one visit to have better signal to noise ratio ( hereafter `` visit stacks '' ) , subtraction of visit stacks from the reference images to search for variabilities , and creating lightcurves from the subtracted images .
we have shown in @xcite how to obtain lightcurves for resolved sources from the pandromeda data .
the major difference of the data used in this work is that our present data - set contains three years of pandromeda data , instead of one year and a few days from the second year data used in @xcite .
the sky tessellation is also different , in order to have the central region of m31 in the center of a skycell ( skycell 045 ) , instead of at the corner of adjacent skycells ( skycell number 065 , 066 , 077 , and 078 ) as in @xcite ; the skycells are larger and overlap in the new tessellation .
the new tessellation is drawn in fig . 1 of @xcite .
we have extended the analysis to 47 skycells , twice as many as the number of skycells used in @xcite .
the skycells we used are 012 - 017 , 022 - 028 , 032 - 038 , 042 - 048 , 052 - 058 , 062 - 068 , 072 - 077 , which cover an area of 7 deg@xmath5 of m31 .
the search of variability is conducted in both @xmath0 and @xmath6 , where we start from the resolved sources in the @xmath0 reference images , and check for variability in both @xmath0 and @xmath6 filters .
in addition , we also use the deep photometric catalog from the local group galaxy survey ( lggs , * ? ? ? * ) .
the lggs utilized the 4-meter kpno telescope to observe the m31 galaxy .
their mosaic ccd camera has a resolution of 0.261 / pixel at the center which decreases to 0.245 / pixel towards the corner .
the field - of - view of the camera is roughly 36 @xmath10 36. m31 was observed between 2000 and 2002 in johnson _ ubvri _ filters with seeing values from 0.8 to 1.4 .
the observations cover ten fields , corresponding to 2.2 deg@xmath5 along the major axis of m31 ( see fig
. 1 of * ? ? ?
the astrometric solution for each frame was derived by matching with the usno - b1.0 catalog @xcite .
the _ psf _
photometry in each filter was obtained with the iraf daophot routine , and calibrated against the lowell 1.1 m data with zero points and color terms .
the final lggs catalog contains 371,781 stars in m31 , with 1 - 2 percent statistic error at 21 mag and 10 percent at 23 mag .
when matching the lggs catalog to the pandromeda catalog , we found an median astrometric difference of 0.36 arcsec . in this work
we utilize the catalog and images delivered by the two micron all sky survey ( 2mass , * ? ? ?
2mass employed two 1.3-meter telescopes located at mt .
hopkins , arizona and cerro tololo , chile to survey the full sky in three near infrared passbands j(1.25@xmath11
m ) , h(1.65@xmath11 m ) , and k@xmath12(2.16@xmath11 m ) simultaneously .
the pixel scale of the 2mass ccds is 2 arcsec / pixel .
2mass observed every patch of the sky with six times 1.3 s integration time .
the raw images were dark subtracted , flat - fielded , sky subtracted , resampled to a 1 arcsec /pixel coordinate grid in a flux - conserving manner , and coadded to generate an atlas image .
the source detection was performed on the atlas images with psf profile - fitting , yielding a sensitivity of 15.8 , 15.1 , and 14.3 mag at 10 @xmath13 level in j , h , and k@xmath12 band , respectively .
the astrometry was calibrated against the tyco-2 reference catalog and yielded an order of 100 mas accuracy for bright sources .
we queried the 2mass point - source catalog for a 3 @xmath10 3 deg@xmath5 area centered at m31 via the nasa / ipac infrared science archive , and retrieved 43,723 sources in this region .
in addition , we also retrieved postage - stamp images from irsa to examine the sources .
we designed our selection algorithm based on the optical and infrared properties of known lbvs .
the first criterion is the optical color . as can be seen in fig .
[ fig.criteria_vbv ] , the known lbvs and lbv candidates presented by @xcite are rather blue in the b - v v.s . v color - magnitude diagram ( cmd ) .
we thus set the criterion that b - v @xmath14 0.5 mag and v brighter than 20 mag to select for optically blue and luminous objects .
this enables us to filter out possible contaminations from foreground stars and unresolved background galaxies .
the next criterion utilizes the optical variability from pandromeda data .
since ob stars also show aperiodic variation at @xmath14 0.1 magnitude level - the so - called @xmath15 cyg variables - and since have accounted for variation @xmath16 0.4 magnitude in their m33 lbv candidates as by similar mechanism , we require lightcurve variations @xmath2 0.4 mag from @xmath0 to secure mid - term lbv variability . loosing this criterion might help to gain more candidates
. we will come to this point in section [ sec.det_eff ] .
sometimes there are a few data points with huge error - bars which deviate from other observations .
such outlier measurements could render a lightcurve with @xmath17 mag @xmath2 0.4 and leads to false detection . to reduce the number of false detection
, we thus require that in a single lightcurve , at least 25 data points vary at 10-@xmath13 level with respect to the mean value from all the measurements .
the last criterion is set by the infrared color .
@xcite have indicated that b[e ] super giants ( b[e]sgs ) also show variabilities similar to lbvs , and pointed out that lbvs and b[e]sgs can be distinguished from the infrared colors as shown in a recent study by using samples from the milky way and the magallanic clouds .
we plot the sample of in fig .
[ fig.jhk ] , as well as the four known lbvs in m31 .
the m31 lbvs follow the distribution of galactic lbvs ; most of them have h - k @xmath14 0.5 .
we thus set a upper h - k limit of 0.5 to select candidate lbvs .
there are seven sources pass the aforementioned criteria .
we inspect their lightcurves individually and select four as candidate lbvs . for the remaining three sources ,
they are all known lbvs ( ae and , m31 var 15 , and m31 var a-1 ) .
the only known lbv that does not pass our selection criteria is af and , due to its variability smaller than the 0.4 mag criterion ( 0.375 ) during the time - span of ps1 .
the number of stars that passed each criterion is listed in table [ tab.crit ] .
there are in total four previously unknown sources which pass all our criteria .
their optical and infrared photometry , as well as their spatial distribution and lightcurves , are shown in table [ tab.cphot ] , fig .
[ fig.spat ] and [ fig.opt ] , respectively .
we collect uv images of known lbv and lbv candidates listed in @xcite and present in fig .
[ fig.nuvpstampm07 ] and [ fig.fuvpstampm07 ] .
the images are taken at near and far uv by _
galex_. most of the lbv and lbv candidates in @xcite are bright sources in nuv , but some of them show faint or no fuv emission .
we search for hst archive images of the known lbvs listed in @xcite in the `` panchromatic hubble andromeda treasury '' program @xcite , and found two of the known lbvs ( af and and m31 var 15 ) have been observed by this program .
their images are shown in fig .
[ fig.pstamphst3 ] .
we show the lightcurves of four known lbvs from @xcite in fig .
[ fig.opt1 ] .
the lightcurves are from our ps1 three year data .
the photometric variation during the three year time - span are listed in table [ tab.knownlbv ] . | we perform a study on the optical and infrared photometric properties of known luminous blue variables ( lbvs ) in m31 using the sample of lbv candidates from the local group galaxy survey @xcite .
we find that m31 lbv candidates show photometric variability ranging from 0.375 to 1.576 magnitudes in @xmath0 during a three year time - span observed by the pan - starrs 1 andromeda survey ( pandromeda ) .
their near - infrared colors also follow the distribution of galactic lbvs as shown by .
we use these features as selection criteria to search for unknown lbv candidates in m31 .
we thus devise a method to search for candidate lbvs using both optical color from the local group galaxy survey and infrared color from two micron all sky survey , as well as photometric variations observed by pandromeda .
we find four sources exhibiting common properties of known lbvs .
these sources also exhibit uv emission as seen from _ galex _
, which is one of the previously adopted method to search for lbv candidates .
the locations of the lbvs are well aligned with m31 spiral arms as seen in the uv light , suggesting they are evolved stars at young age given their high - mass nature .
we compare these candidates with the latest geneva evolutionary tracks , which show that our new m31 lbv candidates are massive evolved stars with an age of 10 to 100 million years . |
electromagnetic interactions have long been used to study both hadronic structure and strong interaction dynamics .
examples include deep inelastic lepton - nucleon scattering , hadronic production of lepton pairs , the production of photons with large transverse momenta , and various photoproduction processes involving the scattering of real or very low mass virtual photons from hadrons .
in particular , the photoproduction of jets with large transverse momenta is calculable in qcd and offers additional complementary information to that obtained from the study of the hadroproduction of jets @xcite . in the photoproduction case
there are contributions where the photon s energy contributes entirely to the hard - scattering subprocess ; these are often referred to collectively as the direct component .
in addition , a real photon can interact via its hadronic substructure .
these contributions comprise the resolved component , a review of which is contained in ref.@xcite .
therefore , the photoproduction of jets allows one to investigate new production mechanisms , probe the hadronic substructure of the photon , and study the conventional hadronic jet production mechanisms which these processes have in common with the hadroproduction case .
two basic approaches are commonly employed for generating predictions for hard - scattering processes . on the one hand
, it is possible to perform calculations for a specific observable in which the integrals over the subprocess variables are done analytically , leaving only the convolutions with the parton distributions to be done numerically . on the other hand , if the subprocess integrations and the parton distribution convolutions are done using monte carlo techniques , it is possible to generate predictions for a variety of different observables simultaneously .
this latter approach is sometimes referred to as a fully differential monte carlo calculation .
several groups have performed next - to - leading - order calculations of jet photoproduction in varying degrees of generality using one or the other of these two approaches . in @xcite , @xcite , and @xcite
subprocesses which involved the photon were kept up to @xmath0 .
thus , the direct component was calculated to next - to - leading - logarithm accuracy while the resolved component was calculated in the leading - logarithm approximation .
the resolved component was calculated to next - to - leading - logarithm accuracy in @xcite , @xcite , @xcite , and @xcite for single inclusive production . for a review see @xcite .
recently , both direct and resolved contributions calculated to next - to - leading - logarithm accuracy have begun to appear @xcite , @xcite , @xcite in a fully differential form .
the purpose of this work is to present a discussion of a calculation which is based on the phase space slicing method using two cutoffs @xcite . both the direct and resolved components are included to next - to - leading - logarithm accuracy .
the result is fully differential and implemented in a monte carlo style program which allows the simultaneous histogramming of many distributions incorporating experimental cuts .
it represents an improvement of earlier results which included the direct component only at nlo @xcite and an elaboration of the very brief results for both components already presented @xcite .
details of the calculational method are presented as well as comparisons with recent data .
some comments on various unsettled issues that arise when comparing with dijet cross sections are also given .
the remainder of the paper is as follows .
the phase space slicing method is reviewed in sec .
numerical results are compared with h1 and zeus data and related physics issues are discussed in sec .
iii , while the conclusions are given in sec .
in this section we describe the calculation of qcd corrections to two - jet production in electron - proton scattering using the phase space slicing method . before discussing the qcd corrections ,
it is necessary to recount the connection between electron - proton and photon - proton scattering . for small photon virtualities
@xmath1 the two are related using the weizscker - williams approximation @xcite wherein one assumes that the incoming electron beam is equivalent to a broad - band photon beam .
the cross section for electron - proton scattering is then given as a convolution of the photon distribution in an electron and the photon - proton cross section @xmath2 the integration limits and the maximum photon virtuality @xmath3 are determined from the ( anti ) tagging conditions of the experiment .
the energy of the photon is related to the energy of the incident electron by @xmath4 .
we used the improved photon distribution in an electron @xcite given by the formula @xmath5 \right\}\ ] ] where @xmath6 is the electron mass and @xmath7 is the electromagnetic coupling . within this approximation , qcd corrections to electron - proton scattering correspond to qcd corrections to photon - proton scattering to which we now turn . in this version of the phase space slicing method @xcite two small cutoffs @xmath8 and
@xmath9 are used to delineate regions of phase space where soft and collinear singularities occur .
let the four vectors of the three - body subprocesses be labeled @xmath10 , and define the mandelstam invariants @xmath11 and @xmath12 .
consider , for example , the @xmath13 subprocess whose matrix element becomes singular as the energy @xmath14 of the final state gluon becomes soft .
define the soft region @xmath15 by @xmath16 and the complementary hard region @xmath17 by @xmath18 , both in the @xmath19 rest frame .
the two - to - three body contribution to the cross section is then decomposed as @xmath20 where @xmath21 is the three - body squared matrix element and @xmath22 is the three - body phase space
. within @xmath15 one sets @xmath23 @xmath24 except in the denominators of the matrix elements and then analytically integrates over the unobserved degrees of freedom in @xmath25 space - time dimensions .
the result , proportional to the leading order cross section , contains double and single poles in @xmath26 , and double and single logarithms in the soft cutoff @xmath8 .
terms of order @xmath8 are neglected .
next , the collinear regions of phase space are defined to be those where any invariant ( @xmath27 or @xmath28 ) becomes smaller in magnitude than @xmath29 .
the hard region is then decomposed into collinear @xmath30 , and non - collinear @xmath31 , regions as follows : @xmath32 within @xmath33 one retains only the leading pole of the vanishing invariant in the squared matrix elements .
exact collinear kinematics are used to define the integration domain of @xmath33 , which is valid so long as @xmath34 .
the integrations over the unobserved degrees of freedom are then performed analytically in @xmath25 space - time dimensions giving a factorized result where single poles in @xmath26 , and single logarithms in both cutoffs @xmath9 and @xmath8 , multiply splitting functions and lower - order squared matrix elements .
terms of order @xmath9 and @xmath8 are neglected .
the soft and final state hard collinear singularities cancel upon addition of the interference of the leading order diagrams with the renormalized one - loop virtual diagrams .
the remaining initial state collinear singularities are factorized and absorbed into the parton distributions .
the integrations over the singularity - free portion of the three - body phase space @xmath35 are performed using standard monte carlo techniques .
when all of the contributions are combined at the histogramming stage the cutoff dependences cancel , provided one looks at a suitably defined infrared - safe observable .
an example will be given below .
the matrix elements squared for all two - to - two and two - to - three parton - parton scattering subprocesses through @xmath36 are from the paper of ellis and sexton @xcite , and those for the photon - parton scattering subprocesses through @xmath0 are from the paper of aurenche _
et al_. @xcite . in order to compare with experimental data
, a suitable jet definition must be chosen . for this study ,
the snowmass algorithm @xcite has been used , supplemented with an @xmath37 constraint @xcite as follows .
let @xmath38 ( @xmath39 ) be the pseudorapidity and @xmath40 ( @xmath41 ) be the azimuthal angle of parton @xmath42 ( @xmath43 ) and then define @xmath44^{1/2}.\ ] ] partons @xmath42 and @xmath43 are merged if @xmath45.\ ] ] note that @xmath46 corresponds to having no @xmath37 constraint .
the 4-vector recombination is of the type @xmath47 other types of jet algorithms may also be used .
using the results of the method described in the previous section we have constructed a program to calculate jet photoproduction cross sections .
the program uses monte carlo integration so it is possible to implement experimental cuts , provided that they are defined in terms of partonic variables . the cteq4 @xcite proton parton distribution set was used throughout , in conjunction with either the grv @xcite or gs @xcite photon set . unless otherwise noted , the grv set was used .
the calculation was performed in the @xmath48 factorization scheme .
therefore , the corresponding @xmath48 distribution sets were used when working at next - to - leading order ( nlo ) . for leading order ( lo ) results ,
the appropriate lo sets were used .
similarly , the two - loop version of the strong coupling @xmath49 was used with matching across quark thresholds for the nlo results , and at lo the one - loop value was used .
the value of @xmath50 was taken from the proton parton distribution set . for future use
we define three sets of beam energy and tagging conditions as follows : 1993 h1 where @xmath51 , @xmath52 , @xmath53 and @xmath54 , 1993 zeus where @xmath51 , @xmath52 , @xmath55 and @xmath56 , and finally 1994 zeus where @xmath57 , @xmath52 , @xmath55 and @xmath58 . as explained above ,
when contributions from the two - body and three - body pieces are combined to form a suitably defined infrared - safe observable the cutoff dependences cancel .
this provides a check on the method and also on its implementation . as stated in sec .
ii , the method requires that @xmath34 .
therefore , we take @xmath59 with @xmath60 and verify that the result is independent of @xmath61 . as an example , the single jet inclusive cross section integrated over @xmath62 and @xmath63 is shown in fig . 1 for the 1994 zeus conditions with @xmath64 and @xmath65 . the two - body ( negative ) and three - body ( positive ) contributions together with their sum are shown as a function of @xmath8 for @xmath66 ( dash line ) and @xmath67 ( solid line ) .
the bottom enlargement shows the sum for @xmath66 ( circles ) and @xmath67 ( diamonds ) relative to @xmath68 ( dot line ) of the actual result ( solid line ) . for this particular observable , the result is sufficiently independent of the cutoffs below @xmath69 .
one of the primary reasons for carrying out a nlo calculation is that of reducing factorization and renormalization scale dependence .
this is illustrated in fig . 2 for the single jet inclusive cross section as a function of @xmath70 integrated over @xmath71 for the 1993 h1 conditions with @xmath64 and @xmath65 .
the factorization and renormalization scales have been set equal to @xmath72 .
the lo ( dash line ) and nlo ( solid line ) results are shown for two different scale choices @xmath73 ( top ) and @xmath74 ( bottom ) where @xmath75 is the maximum transverse energy associated with the relevent two- or three - body weight . the single jet inclusive cross section as a function of @xmath76 integrated over @xmath77 is shown in fig .
3 . in both cases , the nlo calculation shows reduced scale dependence relative to the lo result . as stated in the introduction , it is conventional to describe jet photoproduction in terms of two components .
the direct or point - like component corresponds to the case where the photon participates wholly in the hard - scattering subprocess , whereas the resolved or hadron - like component corresponds to a situation where the photon interacts as if it contained partons . in the direct component there is , then , no remnant beam jet and this leads to a simpler event structure than in the purely hadronic case .
the decomposition is , however , an artifact of viewing the process from the perspective of leading order qcd . at next - to - leading order ,
the two components mix and a complete separation is not possible .
although a strict separation into direct and resolved components is not possible , one can still define kinematic regions where either component dominates .
one such definition @xcite uses the _ observed _ momentum fraction @xmath78 of the parton coming from the photon which is given by @xmath79 here @xmath80 and @xmath81 are those of the two highest transverse energy jets .
the region @xmath82 corresponds primarily to the direct contribution and @xmath83 to the resolved contribution .
it is interesting to see how the two components populate phase space . to this end ,
shown in fig .
4 is the nlo single jet inclusive cross section as a function of @xmath70 integrated over @xmath71 in three different regions @xmath84 ( solid ) , @xmath82 ( dot ) , and @xmath83 ( dash ) for the 1993 zeus conditions with @xmath64 , and @xmath65 .
the cross section dominated by the resolved component has a steeper @xmath80 dependence than that which is dominated by the direct component .
this is due to the extra convolution associated with the photon parton distribution . in other words ,
only a portion of the photon s energy contributes to the hard scattering for the resolved component .
hence , at large @xmath80 values the direct component dominates . in fig . 5 the nlo single jet inclusive cross section is shown as a function of @xmath76 integrated over @xmath85 .
the direct component dominates at negative rapidity whereas the resolved component takes over in the positive direction . in the coordinate system used in the hera experiments
, positive rapidity corresponds to the direction of the proton beam . therefore ,
if the observed jet is in the positive rapidity region , this corresponds to larger @xmath86 values from the proton and smaller ones from the photon .
conversely , if the jet is in the negative rapidity region , larger photon @xmath86 values and smaller proton @xmath86 values are favored . since the contribution dominated by the direct component has @xmath87 , it dominates in the negative rapidity region , as shown . in fig .
6 the dijet cross section is shown as a function of the invariant mass of the two highest @xmath80 jets .
again , the steeper resolved component dominates over the flatter direct component at low @xmath88 .
having discussed the general properties of the predictions , it is of interest to see how they compare with existing data .
the first comparison is to the single jet inclusive cross sections as measured by h1 @xcite .
the single jet inclusive cross section as a function of @xmath70 along with our nlo result ( gs photon : solid lines , grv photon : dash lines ) for the 1993 h1 conditions with @xmath89 , @xmath64 , and @xmath65 is shown in fig
the data and theory have different slopes with the theory falling below the data at low @xmath90 for both photon sets .
the agreement is slightly better for the case where @xmath91 .
this is seen more clearly in fig .
8 which shows the single jet inclusive cross section as a function of @xmath76 for @xmath92 with @xmath93 . in the proton direction
the theory consistently undershoots the data .
the agreement improves as @xmath94 increases .
one possible explanation for the observed discrepency may be the need to increase the gluonic content of the photon which is poorly constrained in this @xmath86 region
. however , this would be premature as underlying event contributions , which were not removed from the data set or included in our calculation , may also provide an explaination .
we next switch to 1993 zeus conditions and compare to the published zeus data @xcite on single jet inclusive cross sections . in fig .
9 the measured single jet inclusive cross section is shown as a function of @xmath70 along with our nlo result ( gs photon : solid lines , grv photon : dash lines ) with @xmath89 , @xmath64 , and @xmath65 . in this case
the agreement with the data is somewhat better than that shown in fig .
this is further illustrated by the comparison with the single jet inclusive cross section as a function of @xmath76 , shown in fig .
the trend , although less pronounced , is the same as with the h1 data .
data for several different types of observables involving two large transverse momentum jets have been published . in the zeus dijet analysis a sample of dijet events
is obtained by requiring @xmath95 . in principle , this presents a problem in the following sense .
the pair of jets can be accompanied by additional soft gluons .
the singularities associated with the soft gluon emission cancel infrared singularities coming from one - loop contributions .
this cancellation has already taken place in the present calculation .
that is , the soft gluons with energies up to @xmath96 have already been integrated out .
after the cancellation there are residual @xmath97 terms in the two - body portion of the calculation which cancel against logarithms built up by integrating over the unobserved gluon radiation in the three body contribution .
the requirement that both jets have @xmath98 restricts a portion of this integration and leads to an incomplete cancellation of the @xmath97 dependence .
we have , nevertheless , used this definition for the dijet sample , having found that the cutoff dependence of the results is negligible .
however , the problem still exists , at least in principle .
preliminary zeus data @xcite for the dijet cross section as a function of @xmath99 integrated over @xmath100 are shown in figs .
11 and 12 for @xmath101 with @xmath102 .
the solid curves are our results for the 1994 zeus conditions using the grv photon set , @xmath89 , and @xmath64 .
a value of @xmath103 was used following @xcite where it was found that this value used at the parton level most closely corresponds to the ktclus algorithm used to define the jets .
no systematic cutoff dependence was observed for these results , within the limitations of the monte carlo errors .
the solid curves display the same trend that was observed in the single inclusive plots discussed previously insofar as the direct component comes closer to the data than does the resolved , the data are underestimated in the low @xmath80 region , and the agreement gets better as @xmath94 increases .
this is consistent with there being some residual underlying event contribution to the observed jets which has not been removed during the data analysis .
the problem of defining a sample of two jet events in a manner which is amenable to a next - to - leading order theoretical treatment has been studied in @xcite . in order to avoid a dependence on the theoretical cutoffs used in the course of the calculation
, it was proposed to calculate the dijet cross section by allowing the second jet to have a transverse energy less than @xmath94 if the third unobserved jet is soft , i.e. has a transverse energy of less that @xmath104 ( @xmath105 ) .
this approach slightly expands the available phase space when both jets have transverse energies near @xmath94 and , therefore , avoids the problem of a potential cutoff dependence .
results using this algorithm are shown in figs .
11 and 12 by the dashed lines .
the relaxation of the @xmath106 constraint on the second jet generates additional contributions from the @xmath107 subprocesses so that the dashed curves lie above the solid ones and are in better agreement with the data .
however , the cuts do not correspond to those utilized in the definition of the data sample .
given the unresolved question of the underlying event contributions , it is premature to draw detailed conclusions from comparisons with these data .
angular distributions in dijet production are sensitive to the spin of the exchanged particle .
let @xmath108 be the scattering angle of the two highest @xmath80 jets in their center of mass . in lo ,
the direct component is dominated by @xmath109channel quark exchange which has a characteristic angular dependence given by @xmath110 .
the resolved component , however , is dominated by @xmath109channel gluon exchange and has a @xmath111 angular dependence .
this property is preserved at nlo as demonstrated by the solid lines in fig .
13 where the dijet angular distribution @xmath112 is shown as a function of @xmath113 in the regions @xmath82 ( left ) and @xmath83 ( right ) for the 1994 zeus conditions with @xmath64 , and @xmath65 . also shown in the figure is the zeus @xcite measurement and the lo ( dash line ) result .
to better compare shapes , the distributions are normalized to one at @xmath114 .
the shapes are very insensitive to how one handles the dijet definition issue discussed above .
the excellent agreement between the predictions and the data confirms the theoretically expected behavior .
in this paper , results from a new next - to - leading - order program for jet photoproduction have been presented .
the calculation was performed using the phase space slicing method and implemented in a monte carlo style program .
physical observables are independent of the phase space slicing parameters .
as is the case in other jet production processes , the renormalization and factorization scale dependence is reduced relative to leading - order predictions .
a decomposition into direct and resolved components was also studied using a definition based on the observed momentum fraction of the parton from the photon .
the direct component dominates at large @xmath80 and in the photon direction while the resolved dominates at low @xmath80 and in the proton direction .
comparisons with single inclusive jet cross section measurements by h1 and zeus as a function of @xmath90 and @xmath115 show that the theory tends to lie somewhat below the data in the forward rapidity region .
this is the region corresponding to low values for @xmath78 and which is dominated by the resolved component .
one might suspect that this indicates a need for increasing the size of the photon parton distributions in this @xmath86 range , but this conclusion would be premature since contributions from underlying event structure could account for this , as well .
our results were also compared with dijet cross section measurements from zeus . the cuts used to define the dijet sample are such as to restrict a portion of the phase space needed to insure the proper cancellation of the cutoff dependence in the two- and three - body contributions . in practice
, however , the variation was found to be smaller than the statistical errors associated with the monte carlo integrations .
this point aside , the predictions for the normalized angular distributions agree well with the data .
the @xmath116 distributions show the same pattern of deviations from the theoretical predictions as was observed in the single inclusive jet measurements .
as additional data are acquired it is possible that jet photoproduction will be used to place additional constraints on the parton distributions in the photon .
such data would complement data for the photon structure function which are primarily sensitive to the quark distributions in the photon .
the jet photoproduction process , on the other hand , directly involves the gluon distribution in the photon .
however , in order for such data to provide useful constraints on the photon parton distributions , the effects of the underlying event structure must be minimized .
higher @xmath80 thresholds combined with a smaller cone size in the jet definition would help in this regard .
p. aurenche , l. bourhis , m. fontannaz , and j .- ph .
guillet , in _ proceedings of the workshop 1995/96 on future physics at hera , hamburg , 1996 _ , edited by g. ingelman _
( deutches elektronen - synchrotron , hamburg , 1996 ) , p. 570 .
p. aurenche , r. baier , a. douiri , m. fontannaz , and d. schiff , .
the expressions for the @xmath117 and @xmath118 subprocesses contain a number of typographical errors where the variable @xmath119 appears in place of @xmath120 .
the original report lpthe orsay 86/24 contains the correct expressions .
butterworth , l. feld , m. klasen , and g. kramer , in _ proceedings of the workshop 1995/96 on future physics at hera , hamburg , 1996 _ , edited by g. ingelman _
( deutches elektronen - synchrotron , hamburg , 1996 ) , p. 554 | a new next - to - leading order monte carlo program for the calculation of jet cross sections in photoproduction is described . the contributions from both resolved and direct components are included to @xmath0 .
properties of the predictions for various inclusive jet and dijet observables are discussed and comparisons with hera data are presented . |
recently , the high - energy spectroscopic system ( hess ) collaboration has reported the detection of gamma - rays from the region of the galactic center with energies ranging between approximately 250 gev and 10 tev @xcite .
two other atmospheric cerenkov telescopes ( acts ) , cangaroo - ii @xcite and whipple @xcite have also reported positive detections of very high - energy gamma - rays from the region , although with considerably lower significance .
currently , the origin of this very high - energy emission is unknown .
efforts have been made to explain the data with astrophysical processes involving the supermassive black hole known to exist at the galactic center @xcite . although these are certainly possible explanations , the source of the observed emission is not yet clear .
gamma - ray observations of the galactic center have long been studied as one of the most promising methods to search for particle dark matter @xcite .
the most popular dark matter candidates @xcite , such as the lightest neutralino in @xmath2-parity conserving supersymmetric models , are generally much lighter than the highest energy gamma - rays seen by hess , and therefore can not possibly generate the observed spectrum .
it has been shown that a dark matter particle annihilating through typical channels ( @xmath3 ) would require a mass between 12 and 29 tev to generate this spectrum @xcite .
in addition to the unacceptable levels of fine tuning which would be required , a neutralino in this mass range would generate far more dark matter during thermal freeze - out than is observed .
for these reasons , we do not consider a very heavy neutralino to be a likely source of the flux observed by hess .
models of gauge mediated supersymmetry breaking ( gmsb ) provide an interesting alternative to neutralino ( or other lightest supersymmetric particle ) dark matter .
typically in these models , supersymmetry breaking originates in a strongly interacting sector ( the supersymmetry breaking sector ) and is communicated to the superpartners of the standard model ( sm ) states by sm gauge interactions with particles within a messenger sector @xcite .
it is natural for the lightest messenger particle to be stable , electrically neutral and have a mass on the order of a few times 10 tev .
such a particle , as we will show , can have an annihilation cross section which generates a thermal relic density consistent with the measured dark matter density of the universe . in this letter
, we calculate the gamma - ray spectrum generated in the annihilations of gmsb messenger dark matter and show that this particle can be a viable source of the very high - energy gamma - rays observed by hess .
gmsb models typically contain a messenger sector , charged under the sm gauge group , which communicates supersymmetry breaking to the superpartners of sm states by sm gauge interactions .
such gauge communication of supersymmetry breaking has the great advantage that unobserved flavour - changing - neutral - current , @xmath4-violating , and rare decay processes arising from the soft - supersymmetry - breaking terms are automatically absent , in contrast to most other methods of communication .
since supersymmetry breaking is communicated by renormalizable gauge interactions , the fundamental scale of supersymmetry breaking in gmsb models is much lower than for supergravity communication , possibly as low as 50 to 100 tev depending on the interactions between the superymmetry breaking and messenger sectors . as a consequence ,
the lightest supersymmetric particle ( lsp ) is the gravitino , with mass , @xmath5 ev . such a mass is far too small to contribute substantially to the dark matter density of the universe .
thus , in this class of models , it is necessary to look beyond the lsp for the constituents of dark matter .
fortunately in the simplest class of gmsb models , the messenger sector naturally contains a stable state with mass of a few ten s of tev .
messenger particles in gmsb models have been studied as potential dark matter candidates in the past . in refs .
@xcite , minimal supersymmetric models with gauge mediated supersymmetry breaking were studied . in these models ,
the messenger sector consists only of particles carrying su(3)@xmath6su(2)@xmath6u(1 ) charge , the lightest of which will be stable if the supersymmetry breaking sector contains only singlets charged under this group . in substantial portions of the parameter space
, the lightest messenger state is a massive complex scalar with the sm gauge quantum numbers of a left handed neutrino .
this simple model ultimately fails to provide a viable dark matter candidate , however . with only gauge interactions to contribute to its annihilation rate in the early universe
, it has been shown that such a particle must be lighter than a few tev to not overclose the universe @xcite .
in addition to being too light to produce the very high - energy gamma - rays observed by hess , such a dark matter candidate is excluded by direct detection searches . a very simple way to extend the minimal supersymmetric standard model ( mssm ) is to introduce an additional higgs singlet .
this model , the next - to - minimal supersymmetric standard model ( nmssm ) , has several advantages compared to the mssm .
most notably , it provides a natural mechanism for the @xmath7-parameter to obtain a value near the electroweak scale ( which is required for acceptable electroweak symmetry breaking ) , rather than @xmath8 or @xmath9 .
it also ameliorates the mild fine tuning that now must be imposed on the mssm due to the lep ii constraints on the higgs mass .
the structure of the nmssm can be used to motivate an extension of the gmsb messenger sector to include such an extra higgs singlet field , @xmath10 @xcite .
the superpotential for this sector is then given by : @xmath11 here @xmath12 is a standard model gauge singlet spurion which parameterizes the supersymmetry breaking sector , with expectation values for the scalar and supersymmetry - breaking @xmath13 components , @xmath14 and @xmath15 .
the messenger fields @xmath16 and @xmath17 are in complete @xmath18 and @xmath19 representations of su(5)@xmath20su(3)@xmath6su(2)@xmath6u(1 ) , so contain su(3 ) triplet states with the gauge quantum numbers of the right handed down quarks and su(2)-doublet states with the gauge quantum numbers of the left handed lepton doublet .
the higgs doublets @xmath21 and @xmath22 couple to up and down type fermions , respectively .
the parameters @xmath23 and @xmath24 are expected to be @xmath25 while @xmath26 and @xmath27 have to to be considerably smaller ( @xmath010@xmath28 ) to give a consistent phenomenology for the @xmath7 and @xmath29 parameters @xcite . at the tree level ,
the fermion components of the @xmath30 fields have , as a result of @xmath31 , a common dirac mass , @xmath32 , while because of susy - breaking the scalar components split into two groups with mass - squared @xmath33 . at the one - loop level ,
the masses of the lighter group of scalars are further split , and the lightest of these scalars is the dark matter candidate .
a computation shows that the color - charged scalars are significantly raised in mass , while the splitting between the lightest electrically charged , @xmath34 , and neutral , @xmath35 , messenger is @xcite @xmath36 -\log\left[\frac{m^2+f}{m^2-f}\right ] + \frac{2f}{m^2-f}\log\left[\frac{2f}{m^2+f}\right ] -4 \right ) .
\label{split}\ ] ] a physically acceptable model of messenger dark matter requires that this splitting be positive , so that the neutral scalar is the lightest state , and large enough such that charged state @xmath34 decays via @xmath37 to avoid the presence of exotic quasi - stable charged particles .
this is too restrictive , however , as for messenger masses in the 20 tev range under consideration the total energy deposited from such decays is well below the limits set by bbn photodisociation .
] this requires @xmath38 mev , which for @xmath39 tev imposes @xmath40 , leading to @xmath41 tev , while for @xmath42 tev , @xmath43 is necessary , leading to @xmath44 tev .
thus we see that the physically relevant range of candidate dark matter messenger masses is @xmath45 tev .
the potential that results from the minimizing of the superpotential eqn .
[ superpot ] includes the terms : @xmath46 these three terms represent the vertices @xmath10-@xmath17-@xmath16 , @xmath10-@xmath10-@xmath10 and @xmath10-@xmath21-@xmath22 , respectively , each with a coupling enhanced by the vacuum expectation value of @xmath12 , @xmath47 tev .
it is because of this large vacuum expectation value that other terms in the potential which do not contain a factor of @xmath12 can be safely neglected in our calculation .
furthermore , we can neglect terms that include the small parameters @xmath26 and @xmath27 .
these couplings lead to the following diagrams for messenger annihilation : ( 222,204)(3,4.3 ) ( 0.,20.)(15.,10.)(0.,)/straight+1 ( -5.,18.0)[b]@xmath16 ( 15.,-5.0)(15.,10.0)(0.,)/straight-1 ( 0.,-15.0)(15.,-5.0)(0.,)/straight-1 ( -5.,-18.0)[b]@xmath17 ( 10.,0.0)[b]@xmath16 ( 15.,10.)(30.,20.)(0.,)/scalardash0 ( 15.,-5.)(30.,-15.0)(0.,)/scalardash0 ( 35.,21.5)[t]@xmath10 ( 35.,-15.)[t]@xmath10 ( 222,204)(3,4.3 ) ( 50.,14.)(65.,3.)(0.,)/straight+1 ( 55.,14.0)[b]@xmath16 ( 50.,-7.0)(65.,3.0)(0.,)/straight-1 ( 55.,-10.0)[b]@xmath17 ( 65.,3.)(85.,3.0)(0.,)/scalardash0 ( 75.,10.)[t]@xmath10 ( 100.,14.)(85.,3.0)(0.,)/scalardash0 ( 95.,17.)[t]@xmath10 ( 100.,-7.)(85.,3.0)(0.,)/scalardash0 ( 95.,-7.)[t]@xmath10 ( 222,204)(3,4.3 ) ( 115.,14.)(130.,3.)(0.,)/straight+1 ( 120.,14.0)[b]@xmath16 ( 115.,-7.0)(130.,3.0)(0.,)/straight-1 ( 120.,-10.0)[b]@xmath17 ( 130.,3.)(150.,3.0)(0.,)/scalardash0 ( 140.,10.)[t]@xmath10 ( 165.,14.)(150.,3.0)(0.,)/scalardash0 ( 160.,18.)[t]@xmath21 ( 165.,-7.)(150.,3.0)(0.,)/scalardash0 ( 160.,-7.)[t]@xmath22 the thermally averaged annihilation cross sections corresponding to these diagrams are given by : @xmath48\eta_n^2 k^2 ( 2 \xi_s -\eta_n)^2 - \frac{1}{2}\xi_n^3 ( 2 \xi_s -\eta_n)^3 \eta_n k % % @ \bigg ) \nonumber \\
+ \frac{\langle s\rangle^4}{256 \pi m_{\phi^0}^6 } \bigg(-\frac{9}{2}\xi_n^4 ( 2 \xi_s -\eta_n)^4 - \frac{3}{2}\bigg[1-\frac{1}{\sqrt{2}}\bigg]\eta_n^2 k^2 ( 2 \xi_s -\eta_n)^2 + \frac{11}{4}\xi_n^3 ( 2 \xi_s -\eta_n)^3 \eta_n k % % @ \bigg)\frac{t_{fo}}{m_{\phi^0 } } , \,\,\ , \label{nn}\end{aligned}\ ] ] and @xmath49 where @xmath50 is the temperature at which @xmath51 `` freezes out '' of thermal equalibrium , given by @xmath52 where @xmath9 is the planck mass , @xmath53 is an @xmath25 constant and @xmath54 is the number of degrees of freedom below the freeze out temperature ( @xmath55 for the particle content of the nmssm ) .
when solved by iteration , values of approximately @xmath56 are found . with this in mind , we can see by inspecting eqns .
[ nn ] and [ hh ] that the temperature independent portion of the annihilation cross section ( the first term in each expression ) dominates at the freeze - out temperature .
after freeze - out , the relic density of @xmath35 which remains today is given by @xmath57 thus we conclude that to obtain the cosmologically measured density of dark matter ( @xmath58 ) , we must require @xmath59 ( @xmath60 ) .
assuming @xmath61 and @xmath24 are order one constants , as they are expected to be , and @xmath62 100 tev , we can see from eqns .
[ nn ] and [ hh ] that the required cross section is consistent with @xmath63 . in a parameter space scan of this model ,
the authors of ref .
@xcite have found masses in the range of approximately 7 to 24 tev consistent with @xmath64 .
this range is somewhat artificial , however , depending on the range over which parameters were allowed to vary .
in a previous study of the hess galactic center source @xcite , it was found that the a wimp with a mass in the range of 12 to 29 tev could reproduce the observed spectrum , with the best fit occurring for 19 tev .
this conclusion depends , however , on the assumptions made regarding the annihilation products of the wimp .
the results of ref .
@xcite are therefore not completely general . in the case we are studying here
, the dark matter candidate does not annihilate directly to quark or gauge boson pairs , but rather to pairs of higgs bosons .
although the dominant decays of these higgs bosons depend on the parameters chosen in this sector , in most models , neutral higgses decay to either heavy quarks or gauge bosons . therefore , the gamma - ray spectrum produced in the annihilations of @xmath35 s resembles the spectrum found in ref .
@xcite , but for a wimp with half the mass .
@xmath35 annihilations can also produce pairs of charged higgs bosons .
although the gamma - rays produced in the decays of charged higgses contribute minimally to the overall spectrum , annihilation diagrams with @xmath65 final states can contribute substantially at high energies .
such final states are guaranteed to be present , with the fraction of charged final states that include a radiated photon given by @xmath66,\ ] ] where @xmath67 . as we are considering @xmath68 tev and @xmath69 of a few hundred gev
, this contribution can be significant .
of course this depends on the fraction of @xmath35 annihilations that go to charged higgs pairs . from eqns .
[ nn ] and [ hh ] , we can see that annihilations to either @xmath70 or @xmath71 can dominate , for appropriate choices of parameters . for annihilations to @xmath71 ,
approximately half of which produce @xmath72 final states , we can expect an important contribution from final state photons .
annihilations to @xmath70 do not produce charged higgses , as they mix only into neutral states . in figure [ spec ] , we show the spectrum of gamma - rays predicted for annihilations of @xmath35 s without a charged higgs contribution ( solid ) and with 50@xmath73 of annihilations producing a charged higgs pair ( dashed ) .
we have used @xmath74 gev and @xmath75 gev , although our results depend very weakly on these choices .
the spectra were generated using pythia @xcite as implemented in the darksusy package @xcite .
the gamma - ray spectra predicted from messenger dark matter annihilations can provide good fits to the hess data .
for the spectra shown in figure [ spec ] , with @xmath76 tev , we find fits with @xmath77 = 1.44 ( solid line ) and 1.20 ( dashed line ) per degree of freedom . for a somewhat lighter wimp with @xmath78 tev ,
the quality drops to @xmath77 = 1.75 ( solid line ) and 1.36 ( dashed line ) per degree of freedom . for a 30 tev wimp ,
the @xmath77 s improve to 1.35 and 1.16 per degree of freedom with and without a charged higgs component , respectively .
it is clear that the presence of final state photons from charged higgs diagrams improves the fit to the hess data considerably . in all cases ,
the overall normalization of the spectrum was selected to provide the best possible fit to the hess data . to normalize the gamma - ray spectrum from @xmath35 annihilations ,
the distribution of dark matter near the galactic center must be considered .
unfortunately , the central region of the galaxy is dominated by baryons , so observations of rotation curves reveal little about the distribution of dark matter in this region .
furthermore , simulations have limited resolution and have difficulties predicting the structure of dark matter halos below the kiloparsec scale . with these reservations in mind
, we will discuss some possibilities for the dark matter distribution in the central region of our galaxy .
perhaps the most popular class of distributions are those found by groups using n - body simulations .
these include the well known navarro frenk and white ( nfw ) @xcite and moore _
@xcite profiles .
these distributions display a dark matter cusp in the central kiloparsecs of the galaxy with the behavior @xmath79 with @xmath80 and 1.5 for the nfw and moore _
_ profiles , respectively .
these n - body simulations do not include the effects of baryons on the dark matter distribution , however .
in particular , it has recently been argued that the cooling of baryons in the inner kiloparsecs of our galaxy will compress the dark matter distribution , resulting in considerably higher densities of dark matter @xcite .
the conclusions drawn regarding this process , called adiabatic compression , are still somewhat controversial , however .
thirdly , the presence of a @xmath81 solar mass black hole present at the dynamical center of our galaxy has lead some to argue that the accretion of dark matter onto this body could generate a density spike with @xmath82 @xcite .
others have challenged this conclusion , however @xcite . the gamma - ray flux produced in the annihilations of dark matter particles near the galactic center
is given by @xmath83 where @xmath84 is the angle from the galactic center , @xmath85 is the wimp s annihilation cross section in the non - relativistic limit , @xmath86 is the differential spectrum of gamma - rays produced in each annihilation and @xmath87 is the dark matter density at a distance , @xmath88 , from the galactic center .
the integral is performed over the line - of - sight of the observation .
we can break this expression into two components , one of which depends only on the properties of the dark matter particle and another which depends only on the dark matter distribution .
the second of these factors is given by @xmath89 we can then write @xmath90 where @xmath91 is the average of @xmath92 over the solid angle @xmath93 ( centered on @xmath94 ) . considering the range of masses and annihilation modes discussed in this letter , and a solid angle consistent with the upper limit reported by hess ( @xmath95 steradians ) , we estimate approximately @xmath96 to be the value of @xmath97 required to generate the observed flux .
the profiles favored by n - body simulations predict considerably smaller values , with @xmath98 and @xmath99 for the nfw and moore _
profiles , respectively . these could potentially be increased by a factor of 10 to 100 due to adiabatic compression , however , reaching the desired range .
the presence of a dark matter density spike from adiabatic accretion onto the central supermassive black hole could also potentially generate such large values .
the origin of the recently observed multi - tev gamma - rays from the galactic center region is not presently understood .
although astrophysical possibilities have been proposed , annihilating dark matter is another possible explanation .
the most popular dark matter candidates , such as the lightest neutralino in models of supersymmetry , are far too light too account for the very high - energy spectrum observed . in this letter , we have considered the possibility that stable particles from the messenger sector of gauge mediated supersymmetry breaking ( gmsb ) models are the major component of the dark matter of the universe and that the multi - tev gamma - ray spectrum observed by hess is the product of their annihilations .
we find that 20 to 30 tev messenger dark matter matter particles can reproduce the observed spectrum .
the overall flux can be produced only if the density of dark matter is very large in the inner several parsecs of the galaxy , perhaps as the result of adibatic compression of the halo due to baryon cooling or the adibatic accretion of dark matter onto the central supermassive black hole .
f w. stecker and a. j. tylka , astrophys .
j. 343 , 169 ( 1989 ) ; f. w. stecker , phys .
b 201 , 529 ( 1988 ) ; j. silk and m. srednicki , phys .
* 53 * , 624 ( 1984 ) ; l. bergstrom , j. edsjo and p. ullio , phys .
* 87 * , 251301 ( 2001 ) ; d. hooper and b. l. dingus , phys .
d , in press , arxiv : astro - ph/0210617 ; v. berezinsky , a. bottino and g. mignola , phys .
b * 325 * , 136 ( 1994 ) [ arxiv : hep - ph/9402215 ] ; l. bergstrom , p. ullio and j. h. buckley , astropart .
phys . * 9 * , 137 ( 1998 ) ; d. hooper , i. de la calle perez , j. silk , f. ferrer and s. sarkar , jcap * 0409 * , 002 ( 2004 ) [ arxiv : astro - ph/0404205 ] . for a review of particle dark matter candidates ,
see : g. bertone , d. hooper and j. silk , phys .
, in press , arxiv : hep - ph/0404175 .
see , e.g. , e. poppitz and s. p. trivedi , phys .
d * 55 * , 5508 ( 1997 ) [ arxiv : hep - ph/9609529 ] ; t. hotta , k. i. izawa and t. yanagida , phys .
d * 55 * , 415 ( 1997 ) [ arxiv : hep - ph/9606203 ] ; n. arkani - hamed , j. march - russell and h. murayama , nucl .
b * 509 * , 3 ( 1998 ) [ arxiv : hep - ph/9701286 ] ; s. dimopoulos , g. f. giudice and a. pomarol , phys .
b * 389 * , 37 ( 1996 ) [ arxiv : hep - ph/9607225 ] .
t. han and r. hempfling , phys .
b * 415 * , 161 ( 1997 ) [ arxiv : hep - ph/9708264 ] .
t. han , d. marfatia and r. j. zhang , phys .
d * 61 * , 013007 ( 2000 ) [ arxiv : hep - ph/9906508 ] .
t. sjostrand , p. eden , c. friberg , l. lonnblad , g. miu , s. mrenna and e. norrbin , comput .
commun .
* 135 * , 238 ( 2001 ) [ arxiv : hep - ph/0010017 ] .
p. gondolo , j. edsjo , l. bergstrom , p. ullio and e. a. baltz , arxiv : astro - ph/0012234 ; http://www.physto.se/ edsjo / darksusy/. j. f. navarro , c. s. frenk and s. d. m. white , astrophys .
j. * 462 * , 563 ( 1996 ) [ arxiv : astro - ph/9508025 ] .
b. moore , t. quinn , f. governato , j. stadel and g. lake , mon . not .
310 * , 1147 ( 1999 ) [ arxiv : astro - ph/9903164 ] ; b. moore , f. governato , t. quinn , j. stadel and g. lake , astrophys .
j. * 499 * , l5 ( 1998 ) [ arxiv : astro - ph/9709051 ] .
p. ullio , h. zhao and m. kamionkowski , phys .
d * 64 * , 043504 ( 2001 ) [ arxiv : astro - ph/0101481 ] ; r. aloisio , p. blasi and a. v. olinto , jcap * 0405 * , 007 ( 2004 ) [ arxiv : astro - ph/0402588 ] ; m. milosavljevic and d. merritt , arxiv : astro - ph/0103350 . | recently , the hess telescope has observed a gamma - ray spectrum from the galactic center extending to at least @xmath010 tev .
although it has been suggested that this could be the product of annihilating dark matter particles , the candidates most frequently discussed ( such as neutralinos ) are far too light to account for this flux . in this letter , we consider stable particles from the messenger sector of gauge mediated supersymmetry breaking models as an alternative dark matter candidate .
we find that a 20 to 30 tev messenger state can provide a thermal relic density consistent with the measured dark matter density of the universe and can indeed generate the spectrum observed by hess .
@xmath1 0.5 cm |
in this section we derive the flat band projected hamiltonian , eq .
( 2 ) , by adding interactions to eq . ( 1 ) .
we first use the tight - binding limit to show that flat bands arise from spin - orbit coupled fermions in optical lattices in the absence of interactions . using the tight - binding limit
we transform the interaction to an on - site hubbard interaction .
we then project the interaction into the flat band to derive eq .
we conclude by showing the limits in which the flat band approximation is accurate .
we start by noting that eq .
( 1 ) yields bloch bands that can be accurately fit with a tight - binding model ( for @xmath113 ) : @xmath114 where the nn hopping @xmath24 is tuned to fit the exact band width of the first bloch band of eq .
( 1 ) with @xmath115 and @xmath116 .
@xmath24 is therefore a single - particle hopping derived entirely from optical lattice parameters . here
@xmath117 creates a fermion spinor .
we use eq .
( [ tightbindh ] ) to derive the flat band projected hamiltonian .
( [ tightbindh ] ) can be solved exactly .
the two lowest eigenvalues are given by : @xmath118 fig .
1 plots the lowest bands .
we have tuned the spin - orbit coupling strength @xmath119 and the zeeman field strength @xmath9 to generate flat bands . for both lattice depths
we chose the same band flatness ratio @xcite , @xmath120 , where @xmath105 is the band width and @xmath121 is the separation between the two lowest bands . the unitary matrix that diagonalizes eq .
( [ tightbindh ] ) is : @xmath122 where @xmath123 and @xmath22 .
the eigenstates of eq .
( [ tightbindh ] ) can then be written in terms of the original fermi operators : @xmath124 where @xmath35 and @xmath125 denote operators for lower and upper bloch bands , respectively ( fig . 1 ) .
we can now use the unitary transformation to study the addition of interactions to the non - interacting model within the lowest flat band .
occupancy of just the lowest flat band implies that eq .
( [ tightbindh ] ) acts as an irrelevant constant , to a first approximation .
the relevant term in the model derives from interactions between atoms . in the tight - binding limit
, @xmath4-wave contact interactions thus yield an on - site projected hubbard interaction : @xmath126 where @xmath16 is the on - site repulsion , @xmath127 projects particles into the lowest flat band , @xmath128 creates a fermion at lattice site @xmath129 in the @xmath130 spin state , and the constant is the zero - point energy of the flat band . by using eq .
( [ utransform ] ) in combination with a fourier transform , eq .
( [ hinteraction ] ) becomes : @xmath131 where @xmath132 is defined in the main text .
this shows that eq .
( 2 ) , follows from the on - site interactions operating in a flat spin - orbit band .
we now argue that all fermions occupy the lowest band in realistic parameter regimes .
this is a valid assumption provided the system only partially fills the lowest band and @xmath121 is greater than inter - band interaction matrix elements .
the @xmath4-wave interaction can be tuned far from a feshbach resonance to ensure that @xmath121 is larger than characteristic interaction energies .
this maintains the flat band condition provided @xmath133 .
we note that , in a perfectly flat band @xmath134 , the problem remains strongly correlated even for small @xmath16 because the interaction becomes the only term in the hamiltonian .
the main text shows that @xmath106 is satisfied for @xmath100k . in this section
we show that application of a small staggered chemical potential spontaneously selects one of the degenerate wigner crystal states .
the lowest energies from our exact diagonalization study of @xmath48 [ eq .
( 2 ) ] are 2-fold degenerate at total momentum @xmath54 .
each state is uniform in the absence of symmetry breaking .
if the ground state is truly uniform , a small perturbation ( much smaller than the gap ) should leave the ground state density intact .
if , however , the ground state prefers to spontaneously select an inhomogeneous configuration , a small non - uniform perturbation should drive the system into one of the crystal configurations . in our case ( eqs . 2 or 3 )
a logical choice for degenerate crystal configurations has one particle at every other site in real space .
the crystals are defined as linear combinations of the two lowest energy states at @xmath54 .
to identify crystalline order we perturb the ground state with a small staggered chemical potential term @xmath55 to check if the ground state tends to spontaneously choose one of the two degenerate wigner crystal states .
( this is similar to methods employed in the numerical determination of staggered magnetization in the antiferromagnetic ising model , where a small staggered magnetic field is also applied to pick one of the two staggered magnetization directions . )
care needs to be taken in the limiting procedure . for a fixed small positive @xmath135
, one needs to extrapolate first the lattice size , i.e. , @xmath136 .
the ground state selected as @xmath137 then denotes one of the two broken - symmetry wigner crystal states .
the other wigner crystal state can be detected in the limit @xmath138 .
this shows that spontaneous breaking of the discrete sublattice symmetry inherent in wigner crystals can be detected with the application of a small staggered chemical potential . in this section
we prove the formulas for the emergent luttinger parameters @xmath87 and @xmath88 in the main text , eqs .
( 6 ) and ( 7 ) . to do this
we apply luttinger liquid theory to the effective model in the main text .
we first bosonize the non - interacting part of the model and then the interacting part .
the effective model [ eq . ( 4 ) ] is : @xmath139 where the non - interacting part can be rewritten in terms of two - component vectors in @xmath18-space : @xmath140 here the prime indicates summation over @xmath141 , @xmath142 is the identity matrix , @xmath143 is the @xmath144-component of the pauli matrix , and @xmath145 .
we diagonalize @xmath146 with a unitary transformation , defined on a restricted momentum range @xmath141 : @xmath147 here @xmath148 and @xmath149 label the two bands established by the sublattice dependent hopping in the effective model .
the non - interacting hamiltonian then becomes : @xmath150 where @xmath151 is defined in the main text [ eq .
( 5 ) ] . to begin the bosonization process
, we pass to the continuum limit and expand the field operators into left and right moving fermion fields around the two fermi points : @xmath152 where we have taken into account the opposite slopes in linearizing the effective single - particle energy dispersions of two bands ( see fig . 3 ) .
we can use the left / right ( l / r ) decomposition to bosonize eq .
( [ hochi ] ) using the usual bosonization methods .
we define four bosonic fields , @xmath153 with @xmath154 and @xmath155 , such that : @xmath156 where `` @xmath157 '' indicates normal ordering , @xmath158 , and @xmath159 corresponds to @xmath160 . with these transformations , eq .
( [ noninteractionh ] ) becomes : @xmath161 this form for @xmath146 explicitly reveals the four - component nature of the excitations near the fermi points : two bands ( @xmath154 ) and two directions of motion ( @xmath162 ) .
we now apply the bosonization procedure to the interacting term in eq .
( [ hexthubeq ] ) , @xmath86 .
we first fourier transform the interaction , apply the unitary transformation using eq .
( [ chi_transform ] ) , and we finally substitute the fourier transform for @xmath163 into @xmath86 . this gives rise to 256 terms .
most of these terms cancel to yield : @xmath164 where we have made use of the fact that even and odd sites correspond to bands 1 and 2 , respectively .
we are now able to bosonize the interaction term using the same transformations as those used above .
we first note that , since we are working at half filling ( @xmath83 ) , umklapp terms will vanish .
we further define two conjugate fields for the two bands : @xmath165 the total bosonized hamiltonian can then be written in terms of the conjugate fields : @xmath166\end{aligned}\ ] ] where @xmath167 and @xmath168 .
the two matrices @xmath169 and @xmath170 are given by : @xmath171 to find the emergent luttinger parameters we diagonalize the matrices defining the above hamiltonian with the following transformation @xcite : @xmath172 where the transformation matrices satisfy @xmath173^t=[t_{\theta}]^{-1}$ ] .
this guarantees that the new conjugate fields that hybridize the bands into normal modes , @xmath174 and @xmath175 , satisfy the canonical commutation relations .
it can be checked that the above condition is fulfilled by the following choice of transformation matrices : @xmath176 where the unitary matrix @xmath177 diagonalizes the rescaled matrix @xmath178 , i.e. , @xmath179 and the diagonal rescaling matrices @xmath180 and @xmath181 are given by : @xmath182 here @xmath183 and @xmath184 are defined in the main text .
the total hamiltonian defining the effective luttinger liquid theory can then be written in terms of the diagonalized hamiltonian for each of the normal modes
@xmath185 : @xmath186,\ ] ] where the emergent luttinger parameters @xmath187 and @xmath188 are given in the main text , eqs .
( 6 ) and ( 7 ) . | recent ultracold atomic gas experiments implementing synthetic spin - orbit coupling allow access to flatbands that emphasize interactions . we model
spin - orbit coupled fermions in a one - dimensional flat band optical lattice .
we introduce an effective luttinger - liquid theory to show that interactions generate collective excitations with emergent kinetics and fractionalized charge , analogous to properties found in the two - dimensional fractional quantum hall regime .
observation of these excitations would provide an important platform for exploring exotic quantum states derived solely from interactions .
_ introduction .
_ emergent quantum states derived from interactions can exhibit rich structure because they are , by definition , not adiabatically connected to the underlying single - particle states .
two - dimensional ( 2d ) electron gases placed in a strong magnetic field offer seminal examples . in the absence of a magnetic field , 2d
electrons typically demonstrate fermi - liquid behavior , but a strong magnetic field , the fractional quantum hall ( fqh ) limit @xcite , would seem to prevent the formation of a fermi liquid .
this regime is defined by an absence of single - particle kinetic energy that leaves inter - particle interactions to generate many - body quantum states in a flatband ( the lowest landau level ) .
however , it is now well known that interesting properties , such as fractional charge from screening and other kinetic effects @xcite , emerge from interactions in the fqh regime . the remarkable fact that application of an external field first suppresses single - particle properties to leave interactions to generate similar emergent properties leads to a natural question : can these emergent mechanisms manifest in other contexts ? flatbands in one dimension offer a logical analogue @xcite .
the luttinger - liquid paradigm @xcite captures the physics of many one - dimensional ( 1d ) models .
it predicts excitations with , e.g. , fractionalized charge arising from competition between interactions and kinetic energy .
external fields could , in analogy to 2d magnetic fields , be constructed to quench kinetics in 1d , but the absence of kinetics in 1d flatbands would appear to rule out luttinger - liquid behavior . in this letter , we show that kinetics , fractionalized charge excitations , and other luttinger - liquid - like properties emerge solely from interactions in experimentally feasible 1d flatband models .
our proposal relies on recent experimental progress @xcite in engineering synthetic spin - orbit coupling ( soc ) for ultracold atomic gases @xcite .
these experiments show that raman beams can be used to dress atoms with spin - dependent momentums .
rashba ( and/or dresselhaus ) socs governing these dressed states @xcite are tunable to extremes not possible in solids .
recent work shows that rashba coupling in a 1d optical lattice @xcite or gas @xcite can be tuned to yield flatbands , a new limit that could play a role analogous to the lowest landau level @xcite , but interaction effects in a 1d flat rashba soc band remain unexplored .
we study the impact of interactions between two - component fermions in a flat soc band in 1d optical lattices .
we find that the soc elongates single - particle basis states to generate highly non - trivial nearest neighbor ( nn ) interactions @xcite . the extended interactions lead to wigner crystals of spinors with dispersive collective modes .
these excitations are unexpected because they imply kinetics that emerge purely from interactions .
we predict that these excitations also exhibit fractionalized charge even in the flatband limit . to show this
, we must contend with the fact that the absence of single - particle kinetic energy prevents direct application of the luttinger - liquid theory .
we find , instead , that emergent kinetics allows us to introduce an effective luttinger - liquid theory .
we compute the emergent velocities and fractionalized charge of excitations as experimentally verifiable observables .
we also estimate the experimental parameters for observing these excitations .
detection of kinetics and fractionalized charge derived solely from interactions in 1d would have important consequences for the study of emergent luttinger - liquid behavior , in analogy to emergent fractional charge found in the 2d fqh regime . _ model .
_ we consider an equal population of @xmath0 two - component fermions in a 1d optical lattice .
we start with a non - interacting hamiltonian that adds rashba soc to the optical lattice potential @xcite : @xmath1 where @xmath2 is the momentum of particles of mass @xmath3 , @xmath4 is the optical lattice strength , @xmath5 is the optical lattice wave vector , @xmath6 is the recoil energy , @xmath7 is the soc strength , @xmath8 are pauli matrices , and @xmath9 is the zeeman field strength .
we work in units of the lattice spacing , @xmath10 .
figure [ vvsd ] plots the eigenvalues of @xmath11 , @xmath12 to show that eq .
( [ spinorbith ] ) yields flatbands .
we project into the lowest flatband .
projection , achieved by considering only @xmath13 particles , is warranted in the presence of an energy gap between the @xmath13 and @xmath14 bands at low densities @xcite .
we can derive a low - energy hubbard model of interactions operating in such a flat rashba band in the tight binding limit .
in the wannier basis the inter - atom interactions ( e.g. , @xmath4-wave contact interactions between alkali atoms ) become purely on site . after projection to the lowest flatband , the on - site hubbard interaction defines the hamiltonian of the entire system , and is therefore the focus of our study @xcite : @xmath15 where @xmath16 is the on - site hubbard repulsion that defines the only energy scale , @xmath17 creates a fermion at wave vector @xmath18 in the lowest band , and @xmath19 . here
the kronecker delta implies momentum conservation up to a reciprocal lattice vector and @xmath20 is the number of lattice sites .
equation ( [ hflatband ] ) is written in terms of lowest flatband - projected particles using a unitary transformation between the original fermions and flatband fermions , so that @xmath13 particles are defined as spinors of the original atoms @xcite .
we define the unitary transformation in terms of optical lattice parameters : @xmath21/\omega$ ] with @xmath22 , and @xmath23 . here
@xmath24 is the nn hopping @xcite .
of several lowest bloch bands due to soc for @xmath25 ( left ) and @xmath26 ( right ) .
the ratio of the lowest energy gap to the bandwidth is tuned to @xmath27 @xcite in both panels by setting @xmath28 , @xmath29 for @xmath25 , and @xmath30 for @xmath26 .
( inset ) diagonal interaction between @xmath13 particles as a function of inter - site distance , @xmath31 , for @xmath32 and @xmath33 yielding @xmath34.,width=364 ] projection to @xmath13 particles generates non - trivial delocalized single - particle basis states . to see this we fourier transform @xmath35 to real space
the on - site interaction between the original atoms becomes a longer range interaction between @xmath13 particles .
the leading diagonal interaction , @xmath36 , is between nn . here
@xmath37 . the inset of fig .
[ vvsd ] shows the interaction strength , @xmath38 , between @xmath13 particles .
different optical lattice depths lead to the same interaction , where @xmath38 falls off quickly with the dominant interaction given by @xmath39 , provided the band remains flat @xcite .
the inset shows two key results : ( 1 ) the interaction is longer range , and ( 2 ) the form of the interaction is robust over a wide range of @xmath4 . in the following we
can therefore focus on @xmath26 without loss of generality .
equation ( [ hflatband ] ) contains a large number of terms , but by considering a few of the largest terms ( with strengths @xmath40 , @xmath41 , and @xmath42 ) , we argue for intriguing low - energy states .
leading off - diagonal terms in eq .
( [ hflatband ] ) are given by conditional next nearest neighbor ( nnn ) hoppings of @xmath13 particles , i.e. , @xmath43 and @xmath44 , where @xmath45 and @xmath46 .
we note that conditional hopping originates entirely from interactions .
the @xmath40 term is the strongest and should generate crystal states of spinor @xmath13 particles , but conditional hoppings can give rise to emergent kinetics in excitations .
we verify this picture below by combining diagonalization with an effective model . _ numerical results .
_ to more rigorously study eq .
( [ hflatband ] ) , we use numerics to explore the low - energy hilbert space and confine our study to half filling , @xmath47=1/2 .
we note that the absence of kinetic energy excludes the direct use of luttinger - liquid theory .
( scattered symbols ) for the flatband - projected hamiltonian @xmath48 on various lattice sizes , @xmath20 .
the solid curve is the fit of @xmath49 with the effective extended hubbard model [ eq .
( [ hexthub ] ) ] that highlights a dispersive collective mode .
( right panel ) the energy dispersion of a similar classical model [ eq .
( [ dhexthub ] ] with @xmath50 , showing no dispersive collective modes .
the four schematics denote representative ground and excited state configurations of spinor @xmath13 particles ( encircled arrows ) in real space .
, width=336 ] ( scattered symbols ) for the flatband - projected hamiltonian @xmath48 on various lattice sizes , @xmath20 .
the solid curve is the fit of @xmath49 with the effective extended hubbard model [ eq .
( [ hexthub ] ) ] that highlights a dispersive collective mode .
( right panel ) the energy dispersion of a similar classical model [ eq .
( [ dhexthub ] ] with @xmath50 , showing no dispersive collective modes .
the four schematics denote representative ground and excited state configurations of spinor @xmath13 particles ( encircled arrows ) in real space . , width=115 ] ( scattered symbols ) for the flatband - projected hamiltonian @xmath48 on various lattice sizes , @xmath20 .
the solid curve is the fit of @xmath49 with the effective extended hubbard model [ eq . ( [ hexthub ] ) ] that highlights a dispersive collective mode .
( right panel ) the energy dispersion of a similar classical model [ eq .
( [ dhexthub ] ] with @xmath50 , showing no dispersive collective modes .
the four schematics denote representative ground and excited state configurations of spinor @xmath13 particles ( encircled arrows ) in real space .
, width=120 ] ( scattered symbols ) for the flatband - projected hamiltonian @xmath48 on various lattice sizes , @xmath20 .
the solid curve is the fit of @xmath49 with the effective extended hubbard model [ eq .
( [ hexthub ] ) ] that highlights a dispersive collective mode .
( right panel ) the energy dispersion of a similar classical model [ eq .
( [ dhexthub ] ] with @xmath50 , showing no dispersive collective modes .
the four schematics denote representative ground and excited state configurations of spinor @xmath13 particles ( encircled arrows ) in real space .
, width=120 ] ( scattered symbols ) for the flatband - projected hamiltonian @xmath48 on various lattice sizes , @xmath20 .
the solid curve is the fit of @xmath49 with the effective extended hubbard model [ eq . ( [ hexthub ] ) ] that highlights a dispersive collective mode .
( right panel ) the energy dispersion of a similar classical model [ eq .
( [ dhexthub ] ] with @xmath50 , showing no dispersive collective modes .
the four schematics denote representative ground and excited state configurations of spinor @xmath13 particles ( encircled arrows ) in real space .
, width=134 ] we numerically diagonalize eq .
( [ hflatband ] ) using the lanczos algorithm .
translational symmetry allows us to work within a fixed total momentum sector , @xmath51 .
the left panel of fig .
[ fitband ] shows the four lowest total energies , @xmath49 , as a function of the total momentum for several system sizes .
we find data collapse for @xmath52 .
our numerics therefore apply to the thermodynamic limit .
the ground state of @xmath48 is a spinor wigner crystal shown schematically in the left panel of fig .
[ fitband ] , set to @xmath53 .
two wigner crystals ( both with particles at every other site ) are defined in momentum space by a linear combination of wave functions at @xmath54 .
we verify the crystalline nature of the ground state by breaking the degeneracy with a small , staggered chemical potential , @xmath55 , added to eq .
( [ hflatband ] ) . in the @xmath56 limit
, the density shows that the system spontaneously picks one of the two degenerate wigner crystal ground states @xcite .
we have also calculated the charge structure factor @xmath57 .
we find that @xmath58 has well - defined peaks at @xmath59 , indicating wigner crystals .
we , for comparison , numerically solve a diagonal ( classical ) hamiltonian known @xcite to yield wigner crystals : @xmath60 the right panel of fig .
[ fitband ] shows the many - body energy spectrum .
the ground states of @xmath61 coincide with those of eq .
( [ hflatband ] ) , i.e. , at @xmath54 , further showing that the ground states of eq .
( [ hflatband ] ) are classical wigner crystals of spinor @xmath13 particles .
the first excited state of @xmath61 , however , is non - dispersive and lies at an energy @xmath40 .
this is the energy cost of moving one particle in the wigner crystal to a nn site ( see the schematic of this classical excitation in fig .
[ fitband ] , right panel ) .
a comparison of the left and right panels shows that while the ground states are essentially the same , the excited states of eq .
( [ hflatband ] ) are fundamentally different from those of eq .
( [ dhexthub ] ) .
the excited states of eq .
( [ hflatband ] ) , the left panel of fig .
[ fitband ] , exhibit a gap @xmath62 above the ground state .
the conditional hopping terms cause the otherwise degenerate excited band to form a dispersive collective mode .
the off - diagonal conditional hopping terms superpose the classical configurations of @xmath13 particles .
( see the schematic , left panel of fig .
[ fitband ] ) . to better understand the nature of the excited states , we construct an effective model . .fitting parameters ( @xmath63 and @xmath64 ) , the resulting energy differences ( @xmath65 ) , and the wave - function overlaps between @xmath66 and @xmath48 for an @xmath67 system for the ground state at each total momentum sector , @xmath51 , with @xmath68 . here
the small energy differences and high wave - function overlaps indicate the quality of the effective model in capturing the essential physics of the original model . [ cols="^,^,^,^,^",options="header " , ]
_ effective luttinger - liquid theory .
_ we construct an effective model of eq .
( [ hflatband ] ) by adding hopping terms to eq .
( [ dhexthub ] ) .
the effective hopping terms are emergent because they represent kinetics not present in the original model [ eq .
( [ hflatband ] ) ] .
we verify the accuracy of the effective model by comparing energetics and by taking wave - function overlaps .
the effective model is then studied using luttinger - liquid theory on the emergent degrees of freedom .
we capture the effects of conditional hopping with ordinary single - particle hopping terms in an effective extended hubbard model : @xmath69(\chi_i^{\dagger}\chi_{i+2}^{\vphantom{\dagger}}+h.c.)+h_{d } , \label{hexthub}\ ] ] where @xmath70 and @xmath71 are fitting parameters quantifying emergent nnn hopping .
figure [ eklinearlization ] illustrates @xmath70 and @xmath71 in real space .
note that @xmath63 and @xmath64 scale with @xmath16 because we added these parameters to capture the properties of excited states generated entirely by interactions in the original hopping - free model , eq .
( [ hflatband ] ) .
we vary @xmath70 and @xmath71 and numerically solve eq .
( [ hexthub ] ) to get the best fit of @xmath49 while maximizing overlap of the corresponding wave functions .
table [ bestfitparameters ] shows representative ( @xmath67 ) fits for the lowest eigenstates .
the energy differences between @xmath66 and @xmath48 are all within @xmath72 , and the wave - function overlaps for the lowest states are all above @xmath73 with almost @xmath74 overlaps at @xmath75 and @xmath76 .
the overlap between the ground states ( @xmath54 ) is above @xmath77 .
we plot the first excited state of eq .
( [ hexthub ] ) from the best fit parameters as the black curve in the left panel of fig .
[ fitband ] for comparison .
the overlap and energetic comparison show that eq .
( [ hexthub ] ) captures the essential properties of eq .
( [ hflatband ] ) at low energies .
we can therefore use eq .
( [ hexthub ] ) as an effective theory to make predictions for experiments .
we now show that eq .
( [ hexthub ] ) exhibits excitations with fractionalized charge quantified by luttinger - liquid theory .
we first diagonalize the hopping terms in eq .
( [ hexthub ] ) @xcite .
the emergent `` single - particle '' energy dispersion has two energy bands @xmath78 : @xmath79\cos(2k ) , \label{singleparticlenergy}\ ] ] with fermi velocities @xmath80\sin(2k_{f})|$ ] . for @xmath81
each dispersion crosses the fermi level at two fermi points @xmath82 ( fig . [ eklinearlization ] ) .
low - energy excitations near the fermi points therefore consist of two left movers and two right movers . ) .
the dashed and dotted lines cross at fermi points , @xmath83 .
linearization is shown as solid straight lines .
differing slopes indicate asymmetric bands , i.e. , @xmath84 .
inset : schematic of hopping terms used in eq .
( [ hexthub]).,width=355 ] ) .
the dashed and dotted lines cross at fermi points , @xmath83 .
linearization is shown as solid straight lines .
differing slopes indicate asymmetric bands , i.e. , @xmath84 .
inset : schematic of hopping terms used in eq .
( [ hexthub]).,width=192 ] we bosonize eq .
( [ hexthub ] ) to study interaction effects .
we linearize the dispersion at the fermi points @xcite , as depicted in fig .
[ eklinearlization ] .
the elementary excitations near @xmath85 are bosonic and , in the absence of the interacting term in eq .
( [ hexthub ] ) , have the charge of the original flat band particles .
we include @xmath86 and find the normal modes of the bosonized hamiltonian using a unitary transformation and rescaling of the bosonic fields @xcite .
the emergent normal mode luttinger parameters , i.e. , velocity , @xmath87 , and the charge fractionalization ratio , @xmath88 , are given by @xcite @xmath89 where @xmath90 denotes the two normal modes , @xmath91 , @xmath92 , and @xmath93/2 $ ] .
for @xmath94 , we have @xmath95 indicating that the charge has fractionalized for this normal mode . to see this we write the effective charge , @xmath96 , in terms of the original charge , @xmath97 , as @xmath98 where @xmath96 can be inferred from particle number conductance @xcite .
@xmath99 found here can be continuously tuned below unity .
this should be contrasted with fractionally charged excitations in the fqh regime , where the fractions are only rational @xcite .
the luttinger - liquid analysis therefore shows that low - energy collective modes of eq .
( [ hexthub ] ) can be thought of as fractionalized quasiparticles moving along a spinor wigner crystal .
this result , while known in standard luttinger - liquid theories @xcite , is surprising here since the single - particle eigenstates of the physical atoms are inert ( flatband ) particles that derive emergent kinetics from interactions . the close connection between eqs .
( [ hexthub ] ) and ( [ hflatband ] ) also indicates that these modes should be experimentally observable . _ experimental requirements and observables .
_ low temperatures and low atomic losses are , in general , difficult requirements for proposals to engineer strongly correlated quantum states with atomic gases .
most proposals require maximizing @xmath16 by tuning a feshbach resonance to enter strongly correlated regimes . however , feshbach resonances contribute to unwanted heating and losses @xcite , particularly in soc atomic gases @xcite . the flat band regime studied here
circumvents the need for strong @xmath16 ( and therefore a feshbach resonance ) because the system is automatically strongly correlated in the absence of kinetic energy .
we can estimate realistic parameters to show that the flatband regime is attainable .
@xmath100k is one of the best candidates for strong soc with low losses @xcite . for @xmath100k in a 1d optical lattice with @xmath26 , @xmath28 , @xmath30 , and a perpendicular confinement of lattice depth @xmath101 , we find @xmath102 , @xmath103 and @xmath104 , where @xmath105 is the bandwidth .
this shows that even bare @xmath4-wave scattering implies a strongly interacting flatband problem with @xmath106 .
note that the last inequality is a very stringent flatband requirement .
an accurate ( but weaker ) requirement assumes the many - body energy gap @xmath107 ( left panel , fig .
[ fitband ] ) to be larger than the single - particle hopping @xmath108 .
this implies that partial filling of the lowest band allows us to treat eq .
( [ spinorbith ] ) as an irrelevant constant for realistic system parameters .
parabolic confinement will compete with the many - body energy gap to diminish the size of the wigner crystal near the trap center
. the central crystal will give way to edge states when the parabolic trapping potential energy reaches the gap , i.e. , @xmath109 .
the crystal will then be as large as @xmath110 sites .
trapping potentials therefore place a lower bound on the size of the energy gap ( and therefore @xmath16 ) . for @xmath100k we find that even the bare @xmath4-wave scattering length allows significant crystal sizes , @xmath111 sites for realistic trapping strengths , @xmath112hz .
larger interaction strengths will increase the size of the crystal .
observations of the states proposed here are in principle possible with currently available methods .
the spinor wigner crystal state manifests as a peak in the static structure factor of the original fermions , observable with demonstrated probes : noise correlations @xcite or atomic matter wave scattering @xcite .
luttinger - liquid parameters have also been observed by interfering bose - einstein condensates @xcite . detecting fractionalized charge
is more challenging . in the current context
, it could be measured by , e.g. , detection of partial backscattering from an impurity @xcite , optical methods @xcite , or charge pumping @xcite .
_ summary .
_ we predict a set of intriguing collective states of matter in experiments with atomic fermi gases confined to 1d optical lattices and in the presence of synthetic soc .
we constructed and studied a model where the atomic interactions operate in a flatband .
we found that the single - particle basis states are delocalized spinors .
our analysis predicts that flatband spinor particles have surprising properties generated by on - site interactions among the original atoms : nn interactions and effective nnn hopping .
the many - body ground state was found to be a wigner crystal of spinors .
we find that an effective luttinger - liquid theory parametrizes emergent kinetics and fractionalized charge [ eq .
( [ luttingerparameters ] ) ] in the low - energy collective modes , in direct analogy to the mechanism of emergence found in the fqh regime .
we acknowledge helpful comments from i. spielman and support from the aro ( w911nf-12 - 1 - 0335,w911nf-12 - 1 - 0334 ) , afosr ( fa9550 - 11 - 1 - 0313 ) , and darpa - yfa .
99 d. c. tsui , h. l. stormer , and a. c. gossard , phys .
rev .
lett . * 48 * , 1559 ( 1982 ) .
r. b. laughlin , phys . rev . lett . * 50 * , 1395 ( 1983 ) .
s. d. sarma and a. pinczuk , _ perspectives in quantum hall effects _
( john wiley & sons , inc .
, new york , 1997 ) ; j. k. jain , _ composite fermions _ ( cambridge university press , cambridge , 2007 ) .
v. j. goldman and b. su , science * 267 * , 1010 ( 1995 ) ; r. de - picciotto _ et al .
_ , nature ( london ) * 389 * , 162 ( 1997 ) ; l. saminadayar , d. c. glattli , y. jin , and b. etienne , phys . rev .
lett . * 79 * , 2526 ( 1997 ) .
e. westerberg and t. h. hansson , phys .
rev .
b * 47 * , 16554 ( 1993 ) .
a. seidel and d. h. lee , phys .
rev .
lett . * 97 * , 056804 ( 2006 ) .
s. jansen , e. h. lieb , and r. seiler , phys . status solid ( b ) * 245 * , 439 ( 2008 ) .
m. nakamura , z. y. wang , and e. j. bergholtz , phys .
rev .
lett . * 109 * , 016401 ( 2012 ) .
f. d. m. haldane , j. phys .
c * 14 * , 2585 ( 1981 ) .
m. p. a. fisher and l. i. glazman , in _ mesoscopic electron transport _ , edited by l. l. sohn , l. p. kowenhoven , and g. schon , proceedings of the nato advanced study institute ( 1997 ) .
e. miranda , braz .
j. phys .
* 33 * , 3 ( 2003 )
. t. gimarchy , _ quantum physics in one dimension _ ( oxford university press , new york , 2003 ) .
y .- j .
lin , k. jimenez - garcia , and i. b. spielman , nature ( london ) * 471 * , 83 ( 2011 ) .
p. wang _ et al .
_ , phys .
rev .
lett . * 109 * , 095301 ( 2012 ) .
l. cheuk _
et al . _ ,
phys .
rev .
lett .
* 109 * , 095302 ( 2012 ) .
j. y. zhang _ et al . _ ,
phys .
rev .
lett .
* 109 * , 115301 ( 2012 ) .
z. fu _
et al . _ ,
phys .
rev .
a * 87 * , 053619 ( 2013 ) .
r. a. williams , m. c. beeler , l. j. leblanc , k. jimenez - garcia and i. b. spielman , phys .
rev .
lett . * 111 * , 095301 ( 2013 ) .
i. bloch , j. dalibard , and w. zwerger , rev .
mod .
phys . * 80 * , 885 ( 2008 ) .
j. d. sau , r. sensarma , s. powell , i. b. spielman , and s. das sarma , phys .
rev .
b * 83 * , 140510(r ) ( 2011 ) . v. galitski and i. b. spielman , nature ( london ) * 494 * , 49 ( 2013 ) .
y. zhang and c. zhang , phys .
rev .
a * 87 * , 023611 ( 2013 ) .
x. zhou , y. li , z. cai , and c. wu , j. phys .
b * 46 * , 134001 ( 2013 ) .
b. ramachandhran , h. hu , and h. pu , phys . rev .
a * 87 * , 033627 ( 2013 ) . v. w. scarola , phys . rev .
b * 89 * 115136 ( 2014 ) .
r. a. williams _ et al .
_ , science * 335 * , 314 ( 2012 ) .
see supplemental material for further details .
j. hubbard , phys .
rev .
b * 17 * , 494 ( 1978 ) .
n. sedlmayr , p. korell , and j. sirker , phys .
rev .
b * 88 * , 195113 ( 2013 ) .
j. stenger _
et al . _ ,
phys .
rev .
lett . * 82 * , 2422 ( 1999 ) .
r. wei and e. j. mueller , phys .
rev .
a. * 87 * , 042514 ( 2013 ) .
e. altman , e. demler , and m. d. lukin , phys .
rev .
a * 70 * , 013603 ( 2004 ) ; s. foelling _
et al . _ , nature ( london ) * 434 * , 481 ( 2005 ) ; i. b. spielman , w. d. phillips , and j. v. porto , phys .
rev .
lett . * 98 * , 080404 ( 2007 ) .
b. gadway , d. pertot , j. reeves , and d. schneble , nat . phys .
* 8 * , 544 ( 2012 ) .
s. hofferberth _
et al . _ , nat
. phys . * 4 * , 489 ( 2008 ) .
a. j. daley , p. zoller , and b. trauzettel , phys .
rev .
lett . * 100 * , 110404 ( 2008 ) .
j. ruostekoski , g. v. dunne , and j. javanainen , phys .
rev .
lett . * 88 * , 180401 ( 2002 ) .
j. javanainen and j. ruostekoski , phys .
rev .
lett . * 91 * , 150404 ( 2003 ) .
y. qian , m. gong , and c. zhang , phys .
rev .
a * 84 * , 013608 ( 2011 ) .
l. wang , m. troyer , and x. dai , phys .
rev .
lett .
* 111 * , 026802 ( 2013 ) . |
while many biological processes can be well described with classical mechanics , there has been much interest and debate as to the role of quantum effects in biological systems ranging from photosynthetic energy transfer , to photoinduced isomerization in the vision cycle and avian magnetoreception @xcite .
for example , nuclear quantum effects , such as tunneling and zero - point energy ( zpe ) , have been observed to lead to _ kinetic _ isotope effects of greater than 100 in biological proton and proton - coupled electron transfer processes @xcite .
however , the role of nuclear quantum effects in determining the ground state thermodynamic properties of biological systems , manifesting as _ equilibrium _ isotope effects , has gained significantly less attention @xcite . intermediate and ksi@xmath0inhibitor complex .
schematic depiction of ( a ) the ksi@xmath1intermediate complex during the catalytic cycle ( fig .
s1 ) and ( b ) a complex between ksi@xmath2 and phenol , an inhibitor which acts as an intermediate analog .
both the intermediate and inhibitor are stabilized by a hydrogen bond network in the active site of ksi .
( c ) image of ksi@xmath2 with the tyrosine triad enlarged and the atoms o16 , h16 , o32 , h32 and o57 labeled ( shown with tyr57 deprotonated ) @xcite . ] ketosteroid isomerase ( ksi ) possesses one of the highest enzyme unimolecular rate constants and is thus considered a paradigm of proton transfer catalysis in enzymology @xcite .
ksi s remarkable rate is intimately connected to the formation of a hydrogen bond network in its active site ( fig .
1a ) which acts to stabilize a charged dienolate intermediate , lowering its free energy by @xmath311 kcal / mol relative to solution ( fig .
s1 ) @xcite . this extended hydrogen bond network in the active site links the substrate to asp103 and tyr16 , with the latter further hydrogen bonded to tyr57 and tyr32 as shown in fig .
1a . the mutant ksi@xmath2 preserves the structure of the wild - type enzyme while mimicking the protonation state of residue 40 in the intermediate complex ( fig .
1b ) , therefore permitting experimental investigation of an intermediate - like state of the enzyme @xcite .
experiments have identified that , in the absence of an inhibitor , one of the residues in the active site of ksi@xmath2 is deprotonated @xcite . although one might expect the carboxylic acid of asp103 to be deprotonated
, the combination of recent @xmath4c nmr and uv - vis experiments has shown that the ionization resides primarily on the hydroxyl group of tyr57 , which possesses an anomalously low @xmath5 of 6.3 @xmath6 0.1 @xcite .
such a large tyrosine acidity is often associated with specific stabilizing electrostatic interactions ( such as a metal ion or cationic residue in close proximity ) , which is not the case here suggesting an additional stabilization mechanism is at play @xcite .
one possible explanation is suggested by the close proximity of the oxygen atoms ( o ) on the side chains of the adjacent residues tyr16 ( o16 ) and tyr32 ( o32 ) to the deprotonated o on tyr57 ( o57 , fig .
1c ) @xcite . in several high - resolution crystal structures ,
these distances are found to be around 2.6 @xcite , which are much shorter than those observed in hydrogen bonded liquids such as water , where o o distances are typically around 2.85 .
such short heavy atom distances are only slightly larger than those typically associated with low - barrier hydrogen bonds ( lbhb ) @xcite , where extensive proton sharing is expected to occur between the atoms . in addition , at these short distances the proton s position uncertainty ( de broglie wavelength ) becomes comparable to the o
o distance , indicating that nuclear quantum effects could play an important role in stabilizing the deprotonated residue ( fig 1c ) . in this work
, we demonstrate how nuclear quantum effects determine the properties of protons in ksi@xmath2 s active site hydrogen bond network in the absence and presence of an intermediate analog by combining path integral simulations and isotope effect experiments .
to assess the impact of nuclear quantum effects on the anomalous acidity of tyr57 , we measured the isotope effect on the acid dissociation constant upon substituting hydrogens ( h ) in the hydrogen bond network with deuterium ( d ) .
because tyrosinate absorbs light at 300 nm more intensely than tyrosine , titration curves were generated by recording uv spectra of ksi@xmath2 at different @xmath7 s ( where _ l _ is h or d ) @xcite .
these experiments ( fig .
2 and table s1 ) reveal a change in @xmath5 upon h / d substitution ( @xmath8 ) of 1.1 @xmath6 0.14 for tyrosine in the ksi@xmath2 active site .
this @xmath5 isotope effect is much larger than that observed for tyrosine in solution ( @xmath9 = 0.53 @xmath6 0.08 ) , and is also , to the best of our knowledge , the largest recorded @xmath5 isotope effect @xcite .
changes in static equilibrium properties , such as the @xmath5 , upon isotope substitution arise entirely from the quantum mechanical nature of nuclei .
such a large excess isotope effect , defined as @xmath10 , of 0.57 @xmath6 0.16 thus indicates that the tyrosine triad in the active site of ksi@xmath2 exhibits much larger nuclear quantum effects than those observed for tyrosine in aqueous solution .
given the possible role of nuclear quantum effects in ksi@xmath2 , can one estimate how much the quantum nature of protons changes the acidity of tyr57 compared to a situation in which all the nuclei in the enzyme active site were classical ( @xmath11 ) ?
in the quasi - harmonic limit one can show that the @xmath5 varies as the inverse square root of the particle mass ( m ) @xcite . by using this relation and the experimental @xmath7 values
, we can extrapolate to the classical ( @xmath11 ) limit which yields that the @xmath5 of tyr57 in ksi@xmath2 would be 10.1 @xmath6 0.5 if the hydrogens were classical particles ( fig .
relative to the observed @xmath5 of 6.3 @xmath6 0.1 , this implies nuclear quantum effects lower the @xmath5 or tyr57 by 3.8 @xmath6 0.5 units : an almost four orders of magnitude change in the acid dissociation constant . to provide insights into the molecular origins of the nuclear and electronic quantum effects which stabilize the deprotonated tyr57 residue , we performed simulations of ksi@xmath2 . to treat the electronic structure in the active site we performed _
ab initio _ molecular dynamics ( aimd ) simulations using a qm / mm approach @xcite in which the qm region was treated by density functional theory at the b3lyp - d3 level @xcite ( see methods ) .
these simulations allow for bond breaking and formation as dictated by the instantaneous electronic structure rather than pre - defined bonding rules .
aimd simulations are typically performed treating the nuclei as classical particles . however
, a classical treatment of the nuclei would predict that the @xmath5 would not change upon isotope substitution .
nuclear quantum effects can be exactly included in the static equilibrium properties for a given description of the electronic structure using the path integral formalism of quantum mechanics , which exploits the exact mapping of a system of quantum mechanical particles onto a classical system of ring polymers @xcite .
we combined this formalism with on - the - fly electronic structure calculations and performed _ ab initio _ path integral molecular dynamics ( ai - pimd ) simulations of ksi@xmath2 .
these simulations treat both the nuclear and electronic degrees of freedom quantum mechanically in the active site qm region , and also incorporate the fluctuations of the protein and solvent environment in the mm region .
the simulations consisted of between 47 and 68 qm atoms and more than 52,000 mm atoms describing the rest of the protein and solvent ( table s2 ) .
these simulations , which until recently would have been computationally prohibitive , were made possible by accelerating the pimd convergence using a generalized langevin equation @xcite , utilizing new methods to accelerate the extraction of isotope effects @xcite and exploiting graphical processing units ( gpus ) to perform efficient electronic structure theory evaluations via an interface to the terachem code @xcite .
such a combination yielded almost three orders of magnitude speed - up compared to existing ai - pimd approaches , allowing 1.1 ps / day of simulation to be obtained using 6 nvidia titan gpus .
we have recently shown that ai - pimd simulations using the b3lyp - d3 functional give excellent predictions of isotope effects in water , validating such a combination for the simulation of isotope effects in hydrogen bonded systems @xcite . the excess isotope effect , @xmath12 , obtained from our simulations ( see si methods ic ) was 0.50 @xmath6 0.03 , in excellent agreement with the experimental value of 0.57 @xmath6 0.16 .
the average distance between o57 and the adjacent o16 and o32 atoms obtained in our simulations were 2.56 and 2.57 , with a standard deviation in both cases of 0.09 .
the distribution of distances between o16 and o57 explored in the simulation is shown in the fig .
these average o o distances are slightly smaller than ( and within the margin of error of ) those in the starting crystal structure ( @xmath3 2.6 ) @xcite .
as we will discuss below , the close proximity of the neighboring o16 and o32 groups plays a crucial role in the origins of the observed isotope effect .
figure 3a - c shows snapshots from aimd simulations in which the nuclei are treated classically ( fig .
3a ) or quantum mechanically using the path integral formalism ( ai - pimd , fig . 3b and c ) , while videos of the simulation trajectories are provided in si videos 1 - 3 . for the quantum simulations
the h16 and h32 protons are shown as their full ring polymers , which arise from the path integral quantum mechanics formalism .
the spread of the ring polymer representing each proton is related to its de broglie wavelength ( quantum mechanical position uncertainty ) @xcite .
the uncertainty principle dictates that localization of a quantum mechanical particle increases its quantum kinetic energy .
the protons will thus attempt to delocalize , i.e. spread their ring polymers , to reduce this energetic penalty .
the resulting proton positions in fig .
3b arise from the interplay between the chemical environment , such as the covalent o
h bond , which acts to localize the proton and the quantum kinetic energy penalty that must be paid to confine a quantum particle .
inclusion of nuclear quantum effects thus allows the protons to delocalize between the hydroxyl oxygens to mitigate the quantum kinetic energy penalty ( fig .
3b ) , which is not observed classically ( fig . 3a ) .
confinement of d , which due to its larger mass has a smaller position uncertainty , leads to a much less severe quantum kinetic energy penalty and hence less delocalization ( fig .
3c ) . to characterize the degree of proton delocalization
we define a proton sharing coordinate @xmath13 , where @xmath14 is the distance of proton hx from oxygen atom ox and x=16 or 32 .
hence @xmath15 corresponds to a proton that is equidistant between the oxygen atoms of tyrx and tyr57 , while a positive value indicates proton transfer to tyr57 from tyrx .
figures 3 d - f show the probability distribution along the proton sharing coordinates @xmath16 and @xmath17 for classical nuclei and quantum nuclei for h and d , respectively .
the free energies along @xmath16 and @xmath17 are provided in fig .
s3 . in the classical aimd simulation , h16 and h32
remain bound to their respective oxygens throughout the simulation ( @xmath16 and @xmath17 are negative ) with tyr57 ionized 99.96% of the time ( fig .
3d ) . however , upon including nuclear quantum effects ( ai - pimd simulations ) there is a dramatic increase in the range of values @xmath16 and @xmath17 can explore ( fig .
3e ) . in particular ,
the probability that tyr57 is protonated ( @xmath18 ) increases by about 150-fold for h upon including quantum effects ( fig .
3e ) , with the proton `` hole '' equally shifted onto the adjacent tyr16 or tyr32 residues .
proton transfers between the residues are observed frequently ( si videos 2 and 3 ) with site lifetimes on the order of 60 and 200 fs in the h and d simulations , respectively .
although pimd simulations exactly include nuclear quantum effects for calculating static properties , they do not allow rigorous extraction of time - dependent properties ; nevertheless , they offer a crude way to assess the time - scale of the proton motion .
the frequent transfers observed are also consistent with fig .
3e and f , which show a monotonic decrease in the probability along both @xmath16 and @xmath17 , i.e. although the proton transferred state is lower in probability , the proton transfer process along each of the proton sharing coordinates contains no free energy barrier ( fig .
s3 ) and is thus kinetically fast . as an experimental counterpart
, we used chemical shifts of @xmath4c@xmath19-tyr labeled ksi@xmath2 as a measure of fractional ionization of each tyr residue ( si methods ib ) @xcite .
this analysis yielded values of 79% for the tyr57 ionization for h and 86% for d compared with simulated values of 94.2% and 98.3% ( @xmath6 0.3% ) , respectively .
this represents good quantitative agreement , since the population difference amounts to a difference in the relative free energy between experiment and theory of 0.7 kcal / mol an error which is within the expected accuracy of the electronic structure approach employed .
in addition , the change in the ionization of tyr57 obtained experimentally upon exchanging h for d ( 7% ) is in good agreement with the value predicted from our simulations ( 4.1% ) .
the under - prediction of the isotope effect on fractional ionization from our simulations is in line with the slightly low value of the simulated excess isotope effect , consistent with recent observations that the b3lyp - d3 density functional slightly underestimates the degree of proton sharing , and hence isotope effects , in hydrogen bonded systems upon including nuclear quantum effects @xcite .
the large degree of proton sharing with the deprotonated tyr57 residue upon including nuclear quantum effects can be elucidated by considering the potential energy required , @xmath20 , to move a proton in the ksi tyrosine triad from its energetic minimum to a perfectly shared position between the two tyrosine groups ( @xmath21 ) .
@xmath20 depends strongly on the positions of the residues comprising the triad , and in particular on the separation between the proton donor and acceptor oxygen atoms .
figure 4a shows @xmath20 computed as a function of the distance between o16 and o57 , @xmath22 , for the tyrosine triad in the absence of the protein environment .
removing the protein environment allows us to examine how changes in the triad distances from their positions in the enzyme affects the proton delocalization behavior without introducing steric overlaps with other active site residues ( see si methods i d ) .
figure 4a shows that for the range of oxygen distances observed in the tyrosine triad ( @xmath22 = 2.50 2.65 ) @xmath20 for h16 is 3 - 6 kcal / mol .
this is 6 - 12 times the thermal energy ( @xmath23 t ) available at 300 k , leading to a very low thermal probability of the proton shared state ( lower than @xmath24 ) .
however , upon including nuclear quantum effects the system possesses zpe , which in this system is @xmath34 kcal / mol .
the zpe closely matches @xmath20 and thus floods the potential energy wells along the proton sharing coordinate ( fig .
4b ) allowing facile proton sharing ( fig .
3 ) , i.e. inducing a transition to a lbhb type regime where the protons are quantum mechanically delocalized between the hydrogen bonded heavy atoms @xcite . this leads to qualitatively different behavior of the protons in the active site of ksi@xmath2 : _ from classical hydrogen bonding to quantum delocalization . _
the proton delocalization between the residues allows for ionization to be shared among three tyrosines to stabilize the deprotonation of tyr57 , leading to the large observed @xmath5 shift relative to the value in the classical limit ( fig .
this change in proton behavior gives rise to the large excess isotope effect since an o d stretch possesses a zpe of @xmath33 kcal / mol , which is no longer sufficient to fully flood the potential energy well in the proton sharing coordinate @xmath25 .
as the o o separation is decreased below the values observed in ksi s tyrosine triad , @xmath20 becomes negligible compared to the thermal energy ( @xmath30.6 kcal / mol at 300 k ) .
hence , at very short distances ( @xmath26 2.3 ) thermal fluctuations alone permit extensive proton sharing between the residues and the zpe plays a negligible role in determining the protons positions . thus one would expect a small isotope effect . on the other hand , at bond lengths in excess of 2.7 , @xmath20 becomes so large ( @xmath27 8 kcal / mol , fig .
4a ) that the zpe is not sufficient to flood the barriers , also resulting in a small expected isotope effect @xcite .
the large excess isotope effect in ksi@xmath2 thus arises from the close matching of the zpe and depth of the energetic well ( @xmath20 ) which is highly sensitive to the o o distance .
hence although proton delocalization can occur classically at short o o distances ( @xmath26 2.3 ) , nuclear quantum effects allow this to occur for a much wider range of o o distances ( up to @xmath32.6 ) , making delocalization feasible without incurring the steep steric costs that would be associated with bringing oxygen atoms any closer .
the distances in ksi s active site triad motif thus maximize quantum proton delocalization , which acts to stabilize the deprotonated residue . ) as a function of the hydrogen bond donor - acceptor o
o distance ( @xmath22 ) as compared to the zero - point energy .
( a ) @xmath20 as a function of the oxygen - oxygen distance ( @xmath22 ) between o16 and o57 using the tyrosine triad geometry from a crystal structure ( details in si methods i d ) .
inset figure shows the probability distribution of @xmath22 obtained from the ai - pimd simulation of ksi@xmath2 with ionized tyr57 .
the probabilities are normalized by their maximum values .
( b ) potential energy as a function of the proton transfer coordinate , @xmath25 , for @xmath22 = 2.6 indicating values for the hydrogen and deuterium ( o h and o d ) zero point energies , @xmath28 and @xmath29 .
the position of tyr32 is fixed as the proton h16 is scanned along @xmath22 . ] finally , we considered the role of nuclear quantum effects when an intermediate analog participates in the active site hydrogen bond network .
recent experiments have investigated how the binding of intermediate analogs to ksi@xmath2 affects the sharing of ionization along the extended hydrogen bond network that is formed ( fig .
1b ) @xcite .
these experiments identified that ionization sharing is maximized when phenol , whose solution @xmath5 of 10 equals that of the actual intermediate of ksi ( fig .
1a ) , is bound @xcite .
we thus performed ai - pimd simulations of the ksi@xmath0phenol complex .
the protons were observed to be delocalized across the network ( fig .
5 ) with partial ionizations of tyr57 , tyr16 and phenol calculated to be 18.5% , 56.7% and 22.5% ( @xmath6 0.7% ) , compared with estimates from previous experiments using @xmath4c nmr of 40% , 40% , 20% @xcite . hence simulation and experiment are in good agreement that the ionization is shared almost equally among the three residues , i.e. that there is almost no difference ( @xmath26 2@xmath23 t ) in the free energy upon shifting the ionization among any of the three groups .
therefore , the ability of protons to delocalize within the ksi@xmath2 tyrosine triad , initially found in the enzyme in the absence of the intermediate analog and manifested as a strongly perturbed @xmath5 , extends upon incorporation of the intermediate analog , which shifts the center of the ionization along the network from tyr57 to tyr16 . in both cases ,
proton delocalization acts to share the ionization of a negatively charged group , which suggests that ksi could utilize quantum delocalization in its active site hydrogen bond network to distribute the ionization arising in its intermediate complex ( fig .
1a ) so as to provide energetic stabilization .
in conclusion , ksi@xmath2 exhibits a large equilibrium isotope effect in the acidity of its active site tyrosine residues arising from a highly specialized triad motif , consisting of several short o o distances , whose positions enhance quantum delocalization of protons within the active site hydrogen bond network .
this delocalization manifests in a very large isotope effect and substantial acidity shift .
our simulations , which include electronic quantum effects and exactly treat the quantum nature of the nuclei , show qualitatively and quantitatively different proton behavior compared to conventional simulations in which the nuclei are treated classically , and provide good agreement with experiment . the ability to perform such simulations thus offers the opportunity to investigate in unprecedented detail the plethora of systems
in which short - strong hydrogen bonds occur , where incorporating both nuclear and electronic quantum effects is crucial to understand their functions .
wild - type and d40n ksi from pseudomonas putida were over - expressed in bl-21 a1 cells ( invitrogen ) , isolated by affinity chromatography using a custom - designed deoxycholate - bound column resin , and purified by gel - filtration chromatography ( ge healthcare ) as described previously @xcite .
for @xmath4c nmr experiments , @xmath4c@xmath30-tyrosine was incorporated into ksi according to methods described previously @xcite .
a series of buffers was prepared with a pl between 4 and 10 by weighing portions of a weak acid and its sodium - conjugate base salt , and adding the appropriate form of distilled deionized water ( millipore h@xmath31o , spectra stable isotopes sterile - filtered d@xmath31o ( @xmath2799% @xmath32h ) ) .
buffers were prepared at 40 mm .
tyr57 is solvent accessible , so the tyrosine residues in the active site network are expected to be fully deuterated in d@xmath31o solution .
the following buffer systems were used for the following pl ranges : acetic acid / sodium acetate , 45.25 ; sodium monobasic phosphate / dibasic phosphate , 5.58.25 ; sodium bicarbonate / sodium carbonate , 8.510 .
buffers were stored at room temperature with caps firmly sealed .
after preparation of buffers , pl was recorded using an orion2 star glass electrode ( thermo ) , immediately following calibration with standard buffers at ph 4 , 7 , and 10 . in h@xmath31o ,
the ph of the buffer was taken as the reading on the electrode . in d@xmath31o , the pd of the buffer
was calculated by adding 0.41 to the operational ph * from the electrode reading @xcite .
a series of samples for titration was prepared by combining 60 @xmath33l protein ( 100 @xmath33 m stock in buffer - free l@xmath31o ) , buffer ( 150 @xmath33l of 40 mm stock ) , and extra l@xmath31o .
the final samples were 600 @xmath33l , 10 @xmath33 m protein , 10 mm buffer .
uv - vis measurements were carried out on the samples on a lambda 25 spectrophotometer ( perkin elmer ) , acquiring data from 400 to 200 nm with a 1.0 nm data interval and 960 nm / min scan rate and 1.00 nm slit width . for each measurement
a background was taken to the pure buffer of a given pl , before acquiring on the protein - containing sample .
spectra were recorded in duplicate to control for random detector error .
the spectra were baselined by setting the absorption at 320 nm to zero , and the change in absorption at 300 nm was followed at varying pl s , employing a previous established method @xcite to determine the fractional ionization of a tyrosine - tyrosinate pair .
the error in a@xmath34 from comparing duplicate spectra following baselining was generally between 02% .
for each pl , the average a@xmath34 was calculated and converted to an extinction coefficient ( @xmath35 ) .
the titration experiment was repeated on two independently prepared buffer stocks to control for error in buffer preparation .
ai - pimd and aimd simulations were performed using a qm / mm approach of ksi@xmath2 with tyr57 protonated , ksi@xmath2 with tyr57 ionized , ksi@xmath2 with the intermediate analog bound , and tyrosine in aqueous solution .
the simulations were carried out in the nvt ensemble at 300 k with a time step of 0.5 fs .
the path integral generalized langevin equation ( piglet ) approach was utilized , which allowed results within the statistical error bars to be obtained using only 6 path integral beads to represent each particle @xcite .
the electronic structure in the qm region was evaluated using the b3lyp functional @xcite with dispersion corrections @xcite .
the 6 - 31 g * basis set was used as we found it to produce proton transfer potential energy profiles with a mean absolute error of less than 0.4 kcal / mol for this system ( see fig .
s6 ) . energies and forces in the qm region and the electrostatic interactions between the qm and mm regions were obtained using an mpi interface to the gpu - accelerated terachem package @xcite .
atoms in the mm region were described using the amber03 force field @xcite , and the tip3p water model @xcite .
the simulations were performed using periodic boundary conditions with ewald summation to treat long - range electrostatic interactions .
the energies and forces within the mm region , and the lennard - jones interactions between the qm and mm regions were calculated by mpi calls to the lammps molecular dynamics package @xcite .
the qm region of ksi@xmath2 contained the p - methylene phenol side chains of residues tyr16 , tyr32 and tyr57 ( fig .
s5a ) . for ksi@xmath2 with the intermediate analog , residue asp103 and the bound intermediate analog
were also included in the qm region ( fig .
the qm region of tyrosine in solution contained the side chain of the tyrosine residue and the 41 water molecules within 6.5 of the side chain o
all bonds across the qm / mm interface were capped with hydrogen link atoms in the qm region @xcite .
these capping atoms were constrained to be along the bisected bonds and do not interact with the mm region .
the initial configuration of ksi@xmath2 was obtained from a crystal structure @xcite ( pdb i d 1ogx ) .
for ksi@xmath2 with the intermediate analog , a crystal structure @xcite ( pdb i d 3vgn ) was used with the ligand changed to phenol .
the crystal structures were solvated in tip3p water and energy minimized before performing ai - pimd simulations . the initial configuration for tyrosine in aqueous solution
was obtained by solvating the amino acid in tip3p @xcitewater using the amber03 force field @xcite and equilibrating for 5 ns in the npt ensemble at a temperature of 300 k and pressure of 1 bar .
each system was then equilibrated for 10 ps followed by production runs of 30 ps . to calculate the excess isotope effect , @xmath12 , ( si methods ic ) we used the thermodynamic free energy perturbation ( td - fep ) path integral estimator @xcite . combined with an appropriate choice of the integration variable to smooth the free energy derivatives @xcite
, this allowed us to evaluate the isotope effects in the liquid phase using only a single ai - pimd trajectory .
simulations performed with d substitution showed no change within the statistical error bars reported .
t.e.m . acknowledges support from a terman fellowship , an alfred p. sloan research fellowship , a hellman faculty scholar fund fellowship and stanford university start - up funds .
l.w . acknowledges a postdoctoral fellowship from the stanford center for molecular analysis and design .
this work used the extreme science and engineering discovery environment ( xsede ) , which is supported by national science foundation grant number aci-1053575 ( project number tg - che140013 ) .
s.d.f . would like to thank the nsf predoctoral fellowship program and the stanford bio - x interdisciplinary graduate fellowship for support .
this work was supported in part by a grant from the nih ( grant gm27738 to sgb ) .
the authors are extremely grateful to christine isborn , nathan luehr and todd martinez for their assistance in interfacing our program to the terachem code .
48ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty http://dx.doi.org/10.1038/nphys2474 [ * * , ( ) ] http://dx.doi.org/10.1046/j.1432-1033.2002.03020.x [ * * , ( ) ] link:\doibase 10.1146/annurev - biochem-051710 - 133623 [ * * , ( ) ] link:\doibase 10.1021/ja102004b [ * * , ( ) ] http://www.jbc.org/content/275/52/41100.abstract [ * * , ( ) ] link:\doibase 10.1021/bi026873i [ * * , ( ) ] , http://www.sciencedirect.com/science/article/pii/s0045206804000550 [ * * , ( ) ] link:\doibase 10.1021/bi061752u [ * * , ( ) ] , link:\doibase 10.1021/ja102714u [ * * , ( ) ] , link:\doibase 10.1021/bi101428e [ * * , ( ) ] , link:\doibase 10.1021/bi4000113 [ * * , ( ) ] , link:\doibase 10.1021/bi301491r [ * * , ( ) ] , http://www.pnas.org/content/early/2012/01/10/1111566109.abstract n2 - understanding the electrostatic forces and features within highly heterogeneous , anisotropic , and chemically complex enzyme active sites and their connection to biological catalysis remains a longstanding challenge , in part due to the paucity of incisive experimental probes of electrostatic properties within proteins . to quantitatively assess the landscape of electrostatic fields at discrete locations and orientations within an enzyme active site , we have incorporated site - specific thiocyanate vibrational probes into multiple positions within bacterial ketosteroid isomerase .
a battery of x - ray crystallographic , vibrational stark spectroscopy , and nmr studies revealed electrostatic field heterogeneity of 8 mv / cm between active site probe locations and widely differing sensitivities of discrete probes to common electrostatic perturbations from mutation , ligand binding , and ph changes .
electrostatic calculations based on active site ionization states assigned by literature precedent and computational pka prediction were unable to quantitatively account for the observed vibrational band shifts .
however , electrostatic models of the d40n mutant gave qualitative agreement with the observed vibrational effects when an unusual ionization of an active site tyrosine with a pka near 7 was included .
uv - absorbance and 13c nmr experiments confirmed the presence of a tyrosinate in the active site , in agreement with electrostatic models .
this work provides the most direct measure of the heterogeneous and anisotropic nature of the electrostatic environment within an enzyme active site , and these measurements provide incisive benchmarks for further developing accurate computational models and a foundation for future tests of electrostatics in enzymatic catalysis .
[ * * , ( ) ] http://www.pnas.org/content/110/30/12271.abstract [ * * , ( ) ] http://www.pnas.org/content/110/28/e2552.abstract [ * * , ( ) ] link:\doibase 10.1021/bi401083b [ * * , ( ) ] link:\doibase 10.1021/ja803928 m [ * * , ( ) ] http://www.sciencemag.org/content/264/5167/1887.abstract [ * * , ( ) ] link:\doibase 10.1126/science.7661899 [ * * , ( ) ] , http://www.pnas.org/content/93/16/8220.abstract [ * * , ( ) ] http://books.google.com/books?id=scgx1ru6uxic[__ ] ( , , ) @noop * * , ( ) link:\doibase 10.1021/ct400911 m [ * * , ( ) ] http://www.sciencedirect.com/science/article/pii/0022283676903119 [ * * , ( ) ] link:\doibase 10.1002/jcc.540110605 [ * * , ( ) ] \doibase 10.1002/(sici)1097 - 461x(1996)60:6<1189::aid - qua7>3.0.co;2-w [ * * , ( ) ] link:\doibase 10.1021/ar970218z [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1021/ct9003004 [ * * , ( ) ] link:\doibase 10.1021/ct3006826 [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.4894287 [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.446740 [ * * , ( ) ] @noop * * , ( ) \doibase http://dx.doi.org/10.1063/1.4873352 [ * * , ( ) ] link:\doibase 10.1021/jp210544w [ * * , ( ) ] link:\doibase 10.1021/bi100074s [ * * , ( ) ] link:\doibase 10.1371/journal.pbio.0040099 [ * * , ( ) ] link:\doibase 10.1021/ac60260a013 [ * * , ( ) ] link:\doibase 10.1002/jcc.10349 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) | enzymes utilize protein architectures to create highly specialized structural motifs that can greatly enhance the rates of complex chemical transformations . here
we use experiments , combined with _ ab initio
_ simulations that exactly include nuclear quantum effects , to show that a triad of strongly hydrogen bonded tyrosine residues within the active site of the enzyme ketosteroid isomerase ( ksi ) facilitates quantum proton delocalization .
this delocalization dramatically stabilizes the deprotonation of an active site tyrosine residue , resulting in a very large isotope effect on its acidity .
when an intermediate analog is docked , it is incorporated into the hydrogen bond network , giving rise to extended quantum proton delocalization in the active site .
these results shed light on the role of nuclear quantum effects in the hydrogen bond network that stabilizes the reactive intermediate of ksi , and the behavior of protons in biological systems containing strong hydrogen bonds . |
we present here l and m band results on v4332 sgr .
recent studies of v4332 sgr have shown that it is an interesting object and the present results further support this view .
v4332 sgr erupted in 1994 in a nova - like outburst with an outburst amplitude of 9.5 magnitude in the visible region .
there was only one detailed study of the object during its outburst ( martini et al .
1999 ) which showed that its eruption was different from that of a classical nova or other classes of eruptive variables .
interest in the object has been rekindled because of the recent outburst of v838 mon which drew considerable attention due to its light - echo ( munari et al .
2002 ; bond et al .
it is believed that v838 mon , v4332 sgr and m31 rv ( a red - variable which exploded in m31 in 1988 ; rich et al .
1989 ) could be members of a new class of eruptive objects ( munari et al .
2002 ; bond et al .
2003 , kimeswenger et al .
the cause of the outburst in these objects does not appear to be satisfactorily explained by conventional mechanisms .
thus new theories have been proposed viz .
a scenario involving the merger of main sequence stars ( soker @xmath2 tylenda , 2003 ) and a hypothesis invoking planetary - capture by an expanding star to explain the eruption ( retter @xmath2 marom , 2003 ) .
the present data indicate that the second mechanism could be viable in v4332 sgr .
recent infrared studies of v4332 sgr have detected several bands of alo at a low rotational temperature of 200 - 300k ( banerjee et al .
a considerable change in the spectral energy distribution ( sed ) of the object was seen between 2mass data of 1998 and observations in 2003 indicating the formation of a dust shell between these two epochs ( banerjee et al .
2003 ) . a better estimate of the epoch when the dust actually formed is discussed here .
optical spectroscopy of v4332 sgr in 2003 , showed an interesting spectrum dominated by very strong emission in the resonance lines of ki and nai ( banerjee @xmath2 ashok 2004 ) .
the sed of the star , derived from optical and ir data , indicated a central star with a black - body temperature of 3250k and an ir excess attributed to a dust component at @xmath3 900k ( banerjee @xmath2 ashok 2004 ) .
observations were done using the 3.8 m uk infrared telescope ( ukirt ) .
spectroscopy was done using the ukirt imaging spectrometer ( uist ) , which uses different grisms to cover the 1.4 - 5 micron range .
@xmath4 ( 3.77 @xmath0 m ) and @xmath5 ( 4.68 @xmath0 m ) band photometry - not available earlier for v4332 sgr - was also done using uist .
flat - fielding , spectral calibration and other reduction procedures were done on the same lines as our earlier @xmath6 study of v4332 sgr ( banerjee et al .
2003).the log of the observations and the observed @xmath7 magnitudes of v4332 sgr are given in table 1 .
figure 1 shows the spectrum - the a - x bands of alo in the @xmath8 band , reported earlier ( banerjee et al .
2003 ) are seen prominently in the present spectrum also but are not discussed here . a remarkable feature
- never seen before in a nova - like object - is the deep , solid - state 3.05 @xmath0 m water - ice band formed due to the o - h stretching mode . at very low temperatures , atoms and molecules
can collide and adhere to a dust grain to produce an ice mantle on the surface .
atoms can migrate from one site to another on the mantle to form a molecule - water ice is believed to form this way with h atoms combining with an o atom .
the presence of cold water - ice around v4332 sgr is extremely unexpected since the ejecta of classical novae generally evolve to high temperatures of @xmath3 10@xmath9k ( the coronal phase ) .
following a standard procedure , we have obtained the optical depth plot of the ice feature by fitting a polynomial to the continuum around it ( gibb et al .
the depth of the ice feature below this continuum was found and converted to an optical depth .
the optical depth plot is shown in figure 2 .
the 3.05 @xmath0 m feature was compared with laboratory data for the optical depth of water - ice at different temperatures ( 10 , 30 , 50 , 120 and 160k ) taken from the leiden database for ice analogslab/ ] . from a @xmath10@xmath11 test to the observed and model data , we find that the 50k data gives the smallest value of @xmath10@xmath11 .
the 30k data also gives a comparable value of @xmath10@xmath11 thus suggesting a low temperature of @xmath330 - 50k for the water ice .
an extended red wing between 3.3 - 3.8 @xmath0 m , which is not well fitted by the models , is seen in the observed data .
this extended ice wing is also seen in several water ice detections but the species responsible for it is unidentified ( gibb et al . 2004 ) . from figure 2 ,
the column density of the water - ice @xmath12 was calculated using @xmath12 = @xmath13@xmath14d@xmath15/@xmath16 , where @xmath16 is the the band strength for water - ice with a laboratory measured value of @xmath16 = 20@xmath1710@xmath18 cm molecule@xmath19 .
while carrying out the integration , we have assumed that the missing data points around 2.65 @xmath0 m ( due to atmospheric cutoff ) , are represented by the data points of the 50k laboratory model in that region .
we obtain a value of @xmath13@xmath14d@xmath15 = 362@xmath2027 cm@xmath19 leading to @xmath12 = ( [email protected])@xmath1710@xmath22 @xmath23 - this value may be used in case of future detection of other ices ( co@xmath24 , ch@xmath25oh , ch@xmath26 etc . ) in v4332 sgr to get a better understanding of the ice composition . another rare feature seen in v4332 sgr is the fundamental band ( @xmath15 = 1 - 0 ) of @xmath1co at 4.67 @xmath0 m in emission .
there appear to be only a few other detections of the co fundamental in emission - mostly towards ysos and herbig aebe stars ( e.g. blake @xmath2 boogert 2004 ; pontoppidan et al .
in a few novae , emission in the co first overtone bands has been seen ( rudy et al .
2003 - table 3 therein ; evans et al .
1996 ) but the detection of the fundamental band appears rare ( lynch et al .
1997 ) . the expanded co ro - vibrational spectrum in v4332 sgr is shown in figure 3 .
individual branch lines from p@xmath27 to p@xmath28 and r@xmath29 to r@xmath28 are clearly seen .
the lines are not resolved at their intrinsic width at the observed resolution of 1000 .
a simple model for the @xmath1co emission was computed assuming lte conditions with the level populations proportional to ( 2j+1)e@xmath30 where b and j are the rotational constant for co and the rotational quantum number respectively ; t is the temperature .
the strength of each p or r branch line was then obtained using the transition probabilities for the lines given by goorvitch ( 1994 ) .
the line positions were also taken from goorvitch ( 1994 ) .
the model spectrum thus obtained was convolved with a gaussian instrument function of fwhm 0.046 @xmath0 m ( i.e. a resolution of 1000 ) . from the model
fits , it is difficult to constrain the temperature of the gas too accurately , but a relatively low temperature of @xmath3300 - 400k is suggested . this could possibly be the reason for the absence in v4332 sgr of the first overtone co bands at 2.3 - 2.45@xmath0 m which need a higher temperature for excitation .
these bands have been modeled to have a temperature in the range 2500 - 4500k in novae ( rudy et al .
2003 ; evans et al . 1996 ) .
further , the strong presence of alo bands in v4332 sgr in the same spectral region as the co first overtone bands , makes it difficult to draw definite conclusions on the absence / presence of the latter . a better estimate for the co temperature than that derived here would require modeling based on higher resolution spectra ( e.g. blake @xmath2 boogert 2004 ; pontoppidan et al .
2002 ) .
the detection of water ice at 30 - 50k in v4332 sgr is very intriguing since such a low temperature component is not expected in novae ejecta .
based on the recent comprehensive survey ( gibb et al .
2004 and references therein ) , most water - ice detections are seen towards embedded ysos / protostars .
such objects are known to have circumstellar disks ( csds ) around them .
in addition , water - ice has also been seen in herbig aebe stars and a few t tauri stars ( creech eakman et al . 2002 ; meeus et al .
2001 , gibb et al .
these are all young objects observationally known to possess csds .
furthermore , all the gaseous components of v4332 sgr are seen in emission i.e. the ki / nai , co , alo and tio lines ( banerjee @xmath2 ashok 2004 ; banerjee et al .
these species are at a low temperature ( 200k for alo ; a low excitation temperature is also implied for the ki / nai lines which need only @xmath3 1.5 - 2ev for excitation ) .
if the region in which these species exist were to be a shell surrounding the central 3250k star of v4332 sgr ( banerjee @xmath2 ashok 2004 ) , the relatively cold gas of these species should lead to lines in absorption .
such is not the case . instead
, if the various species are in a disk , their lines can be expected in emission .
furthermore , in the case of a disk , the central star would also be observable since it will be unobscured ( some obscuration could occur depending on the orientation and thickness of the disk ) .
such is the case in v4332 sgr where the continuum from the central star is clearly seen .
additional support for a csd is the infrared excess seen in v4332 sgr ( banerjee @xmath2 ashok 2004 ) .
this ir excess can be attributed to dust in the disk and not in a shell around the star since the dust shell would obscure the visible radiation of the star .
a similar reasoning is used to explain the ir excess in t tauri stars .
it is also relevant to note that many of the known co fundamental detections are in disk - dominated herbig ae stars ( blake @xmath2 boogert 2004 ) .
the considerable width of the ki lines in v4332 sgr was also shown to be consistent with line broadening arising from rotational motion of gas in a disk ( banerjee @xmath2 ashok 2004 ) .
an additional signature that v4332 sgr has pre - existing matter around it at the time of the 1994 outburst comes from the high - resolution h@xmath31 profiles of the object taken shortly after its outburst .
these profiles ( figure 8 of martini et al . 1999 ) show a deep absorption trough at the center which could be caused by absorption by pre - existing matter . therefore , many of the observed characteristics of v4332 sgr support the existence of a csd around the object . however , it is difficult to rule out the possibility of the matter being in clumps instead of a disk .
it may be pointed out that the maximum value of @xmath14 @xmath3 1 of the ice feature ( figure 2 ) , would imply a large extinction of a@xmath32 @xmath3 14 magnitudes towards the central star , had the 3.05@xmath0 m absorption arisen because of cold , intervening matter in the line of sight ( as in embedded ysos in dark clouds ; whittet 1992 ) .
as the central star appears unobscured ( v= 17.52 , banerjee @xmath2 ashok 2004 ) , the 3.05@xmath0 m feature could instead be caused by ice in the csd viewed against the ir flux from dust in the disc leading to the absorption feature .
it was earlier shown that the @xmath6 fluxes of v4332 sgr changed considerably between 2mass observations of may , 1998 ( @xmath33 = 12.1 , @xmath34 = 11.6 , @xmath35 = 10.992 ) and observations of june , 2003 ( @xmath33 = 13.25 , @xmath34 = 11.986 , @xmath35 = 10.023 ; banerjee et al .
while the 2 mass data showed no ir excess and was well modeled by a 3250k blackbody , the 2003 data showed a composite sed of a blackbody component at 900k ( due to newly formed dust ) and a weakening of the 3250k component due to obscuration by dust .
however , it was not possible to establish with accuracy at which stage between 1998 and 2003 the dust formed .
data from the denis survey is now available and this shows that in september , 1999 v4332 sgr had ir magnitudes of @xmath33= 12.46 and @xmath35 = 10.65 ( denis did not observe in @xmath34 band ) .
a comparison between 2mass and denis @xmath36 magnitudes clearly shows the ir flux increasing towards longer wavelengths indicating that dust formation had begun by the denis epoch .
the formation of dust therefore took place at t@xmath37 @xmath3 5.5 years after the outburst of february 1994 .
this is much larger than the typical timescale of 50 - 100 days for dust formation in classical novae .
the difficulty in having a large dust - condensation time in novae is that the density in the ejecta decreases to a low level - due to expansion - that does not favor dust - formation .
the density @xmath38 in the ejecta of novae that form dust is in the range 10@xmath39 to 10@xmath40g@xmath41 ( gehrz 1988 ) . in v4332 sgr
, the electron density n@xmath42 after @xmath3 2 - 3 months from the outburst was estimated to be 10@xmath43 @xmath41 ( martini et al .
1999 ) . a geometric dilution for the expanding ejecta matter may be a reasonable assumption i.e n@xmath42 @xmath31 1/r@xmath11 provided the ejecta expands freely , unhindered by pre - existing matter .
@xmath31 1/r@xmath11 , after a time of 5.5yr , n@xmath42 should have evolved to @xmath3 10@xmath44 @xmath41 giving a value of @xmath38 = 10@xmath18 - 10@xmath45g@xmath41 .
such a low value of @xmath38 appears unfavorable for dust formation .
it therefore appears that dust formation may not have occured per se in the ejecta of the 1994 outburst .
it is likely that the ejecta , after a stage of free expansion , could have impinged on pre - existing matter around v4332 sgr - sufficiently far from the star - thereby heating it and giving rise to the observed emission at present from different species .
this could be one possible scenario to explain the large observed time for the formation of dust . in a connected aspect
, we note that the present @xmath4 , @xmath5 data of figure 1 , shows the continuum peaking at @xmath3 4.4@xmath0 m indicating a lower , dust temperature than the 900k that was inferred earlier .
however , mid / far - ir observations planned on the spitzer telescope should give more definite information on the dust temperature .
more than a hundred stars with planets ( swps ) are known today and in @xmath315@xmath46 of these the planetary companion is unexpectedly close to the host star ( @xmath47 0.05 au ) .
the orbits of such short - period planets around sun - like stars can be unstable because of tidal dissipation and they can be subsequently consumed by the host stars ( sandquist et al . 2002 ; rasio et al .
more evolved stars are also capable of swallowing their planets as they expand ( siess @xmath2 livio 1999 ) .
the viability of planetary ingestion is observationally supported by the enhanced metallicity seen in many swps - and the presence of li isotopes in one of them - suggesting that planetary material has been accreted by the host star in the past ( santos et al .
2000 ; gonzalez et al . 2001 ; israelian et al .
2003 ) . in particular context of v4332 sgr type of objects , retter and marom ( 2003 ) have explained the multi - peak light curve of v838 mon as due to an expanding star ingesting its planets .
we have shown that there is considerable evidence that v4332 sgr is surrounded by a cold , dusty disk .
it is precisely in such an environment that planets are believed to be born from the coagulation and accretion of solid ice and dust particles ( lissauer 1993 ) .
therefore , while it may appear speculative , it is not entirely unlikely that a planet existed around v4332 sgr and its infall led to the outburst .
the plausibility of this scenario is strengthened since other conventional mechanisms - such as a thermo - nuclear runaway on a white dwarf surface ( invoked for classical , recurrent , symbiotic novae ) or a final helium shell flash ( invoked for a born - again agb star ) - fail to explain the pre- and post - outburst properties of v4332 sgr or similar objects .
it was shown ( retter @xmath2 marom 2003 ) in the analysis for v838 mon , that the gravitational energy released by a 1m@xmath48 planet falling into a solar - mass star is 4@xmath1710@xmath49l@xmath50 .
the outburst luminosity of v4332 sgr is difficult to estimate as the distance @xmath51 to the object is very uncertain . using an estimate of @xmath51 = 300 pc ( martini et al .
1999 ) we have shown that the star has a quiescent luminosity of 0.3@xmath52@xmath50 ( banerjee et al .
a 9.5 magnitude brightening at outburst would yield an outburst luminosity of @xmath3 2@xmath1710@xmath53l@xmath50 .
this is lower than the predicted energy release for the capture of a 1m@xmath48 planet but could be made consistent by revising the distance ( the current distance estimate to v838 mon has been revised upwards by a factor of nearly 10 from its first estimate ) , invoking a smaller mass for the captured planet or that a large part of the released gravitational energy has gone into expansion of the stellar envelope ( retter @xmath2 marom 2003 ) rather than visible radiation .
however , we feel there are many aspects about the nature of v4332 sgr that are still unclear and there is scope for alternative models , apart from that proposed here , to explain its outburst .
research at the prl is funded by the dept . of space , govt . of india .
we thank the ukirt service program for observation time covering a part of these observations .
ukirt is operated by the jac , hawaii on behalf of the uk pparc .
we are grateful to the referee for several helpful comments .
cccccc + date(ut ) & band & resol- & integration & standard & magni- + & & ution & time(s ) & star & tude + & & & & & + spectroscopy & & & & & + 2003 sept .
5.294 & m & 1000 & 312 & bs 6998 & + 2003 sept .
5.359 & kl & 700 & 360 & bs 6998 & + 2003 sept .
5.328 & short l & 650 & 720 & bs 6969 & + 2004 april 15.606 & hk & 500 & 720 & bs 7038 & + 2004 april 29.642 & long l & 1150 & 320 & bs 6998 & + & & & & & + photometry & & & & & + 2003 sept .
5.262 & l & & 252 & hd 161743 & 7.23 @xmath20 0.025 + 2003 sept .
5.276 & m & & 322 & hd 161743 & 6.10 @xmath20 0.03 + + | l and m band observations of the nova - like variable v4332 sgr are presented .
two significant results are obtained viz .
the unusual detection of water ice at 3.05@xmath0 m and the fundamental band of @xmath1co at 4.67@xmath0 m in emission .
the ice feature is a first detection in a nova - like variable while the co emission is rarely seen in novae .
these results , when considered together with other existing data , imply that v4332 sgr could be a young object surrounded by a circumstellar disc containing gas , dust and ice .
the reason for a nova - like outburst to occur in such a system is unclear .
but since planets are believed to form in such disks , it appears plausible that the enigmatic outburst of v4332 sgr could be due to a planetary infall .
we also give a more reliable estimate for an epoch of dust formation around v4332 sgr which appears to have taken place rather late in 1999 - nearly five years after its outburst . |
in cosmological theories of structure formation , non linear objects form when overdense dark matter perturbations turn around , collapse , and settle into virial equilibrium ( e.g. @xcite and references therein ) .
gas initially collapses together with the dark matter , but eventually encounters nearly stationary material that had already collapsed .
since the gas is falling in at supersonic velocities , it is slowed down by hydrodynamical shocks , and these shocks are thought to heat the gas to the virial temperature of the dark matter halo . in spherically symmetric models , and in the absence of dissipation ,
a single strong gaseous shock occurs at approximately half of the turn around radius @xcite , coinciding with the `` virial radius '' of the dark matter halo .
more realistically , the behavior of the post shock gas depends sensitively on its cooling time @xcite . on galactic scales ( @xmath5 ) and below , and increasingly toward high redshifts ( @xmath6 ) , the gas can cool rapidly and loose its pressure support , and hence continue its infall . on these scales , the existence of large
scale shocks have been recently called into question by models in which the bulk of the infalling gas remains cold , and reaches the central regions of the dark halo before encountering shocks @xcite . on larger scales , however , where cooling times are long , such as for galaxy clusters , the existence of virial shocks remains an unambiguous prediction of cosmological structure formation theories .
dimensional hydrodynamical simulations of cluster formation ( e.g. @xcite ) have confirmed the existence of virial shocks , with strong discontinuities in gas density and temperature . these and subsequent simulations have also revealed that the infall is anisotropic , with gas falling in to the cluster potential along cosmic filaments . as a result ,
the radial location and strength of the shocks varies along different directions .
the virial shocks are a fundamental ingredient of cosmic structure formation , and may be responsible for diverse phenomenae , such as generating large scale cosmic magnetic fields @xcite and accelerating electrons to contribute to the diffuse cosmic gamma ray background @xcite . the radial location of the shocks , in principle
, also contains information on the cosmological parameters @xcite . despite their importance
, direct evidence for the existence of such shocks does not yet exist .
the major difficulty in observing the virial shock is that it is expected to lie several mpc ( and several arcminutes ) away from the cluster center , a location at which signals such as the x ray surface brightness @xcite , or galaxy number density / peculiar velocities ( which could reveal density caustics , @xcite ) diminish rapidly . in this paper
, we consider the detectability of virial shocks in future observations of galaxy clusters through the sunyaev - zeldovich ( sz ) effect .
the thermal sz effect is a secondary distortion of the cosmic microwave background ( cmb ) spectrum caused by the hot intra
cluster gas along the line of sight to the surface of last scattering ( see @xcite for a review ) .
the cool cmb photons undergo inverse compton scattering on the hot electrons , gaining on average a small amount of energy in the process , creating an intensity decrement at low frequencies ( @xmath7 ) and an increment at high frequencies .
the sz effect is the dominant source of cmb anisotropy at small angular scales .
the sz effect has recently become a valuable observational tool @xcite .
several programs have begun to map out massive clusters of galaxies , study the intracluster medium ( icm ) , and constrain cosmological parameters .
current instruments are now detecting and imaging clusters at high signal - to - noise , and the next generation of instruments should be capable of mapping significant portions of the sky as a means of finding clusters of galaxies ( see @xcite for a review ) .
several studies have predicted the number of clusters that could be detected in future sz surveys @xcite .
the survey yields are quite impressive .
next generation instruments , such as the atacama cosmology telescope ( act ) , south pole telescope ( spt ) , and the _ planck _ satellite , are expected to detect several clusters per day ; the large resulting samples can be used to select the most massive and most regular clusters that will be best suited for the studies proposed here .
the sz effect is ideally suited to study the `` outskirts '' of clusters , because the sz temperature decrement profile is relatively flat ( e.g. @xmath8 , whereas the x ray emission is proportional to the square of the local density ; @xcite ) .
although our main focus is to assess the significance at which the shocks can be detected , we also consider the more general problem of constraining the cluster gas distribution , as well as the structure of the dark matter halos themselves .
the detection of sharp features , such as the virial shocks , calls for high sensitivity , high resolution maps of the sz surface brightness profile of the most massive clusters . for this reason
, we here focus on predictions appropriate for the atacama large millimeter array ( alma ) , a telescope array expected to be operational in 2012 and deliver @xmath9arcsecond resolution , high sensitivity imaging of clusters .
our results can be scaled to apply to other instruments with different parameters .
this paper is organized as follows . in [ sec : alma ] , we summarize the relevant characteristics of alma . in [ sec : rhoprofiles ] , we describe our models for the structure of galaxy clusters .
these models are based on standard descriptions of gas in hydrostatic equilibrium with a dark matter halo , except that we introduce additional free parameters that allow us to vary the location and sharpness of the virial shock . in [ sec : szprofiles ] , we compute and contrast the sz surface brightness profiles in models with different virial shocks . in [ sec : noise ] , we discuss the sources of noise in the sz surface brightness maps . in [ sec : edgedetection ] , we present simple estimates to argue that , in the face of noise , smooth cluster profiles can be distinguished at high significance from profiles that include a virial shock . in [ sec : parameters ] , we go one step further , and compute the statistical accuracy at which the location and sharpness of the shock fronts , as well as other parameters describing the cluster profile , can be determined from future sz maps . finally , in [ sec : conclusions ] we summarize our conclusions and discuss the implications of this work .
several technical points , related to our statistical analysis , are discussed in the appendices . throughout this paper
we assume a standard cold dark matter cosmology ( @xmath10cdm ) , with ( @xmath11 , @xmath12 , @xmath13 , @xmath14 ) = ( 0.7 , 0.3 , 0.045 , 70 km s@xmath15 mpc@xmath15 ) , consistent with the recent results from _ wmap _ @xcite .
the three crucial characteristics of any instrument for the detection of the virial shocks are angular resolution ( needed since the features caused by the shocks are sharp ) , angular coverage ( needed since the features can extend coherently over arcminute scales ) , and sensitivity ( the surface brightness change across the shock front is as small as a few @xmath16k ) .
alma has an ideal combination of these three qualities , and is well suited to studying the gas density cutoff .
alma will be comprised of sixty four 12-meter sub - millimeter quality antennae , with baselines extending from 150 m up to 10 km .
its receivers will cover the range from @xmath17 to @xmath18 ghz .
anticipated sz temperature sensitivity for a @xmath19 integration period is between @xmath20 and @xmath21mk for the compact configuration depending on the frequency of the detection .
the expected aperture efficiency of an antenna has been taken from @xcite .
generally , a baseline @xmath22 corresponds to a beam radius @xmath23 the resulting temperature sensitivities for the compact configuration ( @xmath24 ) is given in table [ tab : alma ] . in our calculations below , we will assume a frequency of 100 ghz , but table [ tab : alma ] shows the required integration time and baseline for several different frequencies .
the second and third columns show the beam diameter and the sensitivity at a fixed integration time of 60s . in the fourth and fifth columns
, we calculated the baseline needed to have @xmath25 and the corresponding sensitivity at the same fixed integration time of @xmath26 . in the last column we calculated the integration time @xmath27 needed to reduce the r.m.s .
noise variance to @xmath1 ( using the scaling @xmath28 ) .
cccccc 35 & 11.79 & 0.1079 & 884 & 3.747 & 67 + 90 & 4.584 & 0.1871 & 344 & 0.983 & 31 + 140 & 2.947 & 0.2518 & 221 & 0.547 & 23 + 230 & 1.794 & 0.4317 & 150 & 0.432 & 31 + 345 & 1.196 & 1.007 & 150 & 1.007 & 169 + 409 & 1.009 & 1.799 & 150 & 1.799 & 539 + 650 & 0.6347 & 13.67 & 150 & 13.67 & @xmath29 + 850 & 0.4853 & 24.46 & 150 & 24.46 & @xmath30 + for the present study we will assume that the sensitivity of the detector is @xmath31k , and its angular resolution is @xmath32 . as table [ tab :
alma ] shows , these assumptions are realistic for alma , requiring @xmath33 hr integration for all frequencies below @xmath34 . for even longer observation times , the detector noise can be neglected compared to the statistical fluctuations of astrophysical origin ( the latter , which has a magnitude of @xmath35k , will be discussed in
[ sec : noise ] below ) .
in this section , we describe the models we adopt for the density and temperature profiles of gas in the cluster .
these models assume that clusters are described by spherically symmetric dark matter halos .
we first follow @xcite to obtain self similar gas density and temperature profiles .
these profiles are then truncated and normalized according to the assumed location and sharpness of a virial shock .
many high - resolution n - body simulations suggest that the dark matter density profile @xmath36 is well described by a self similar form : @xmath37 where @xmath38 is the mass density normalization factor , @xmath39 is a length scale , and @xmath40 is a non - dimensional function representing the profile .
the density normalization , @xmath38 , is determined to yield mass @xmath41 when @xmath36 is integrated within the virial radius , @xmath42 .
the length scale @xmath39 is defined as @xmath43 , where @xmath44 is the concentration parameter .
we use a fitting formula from detailed gas dynamical simulations for the concentration parameter ( see eq .
[ eq : c ] below ) , and calculate @xmath45 from the top
hat collapse model ( see discussion below ) .
the dark - matter profile is approximated by the following analytic form @xmath46 where the parameter @xmath47 is assumed to be either @xmath48 or @xmath49 ( see @xcite , @xcite , and @xcite ) .
the gas density profile can be obtained using a polytropic model in hydrostatic equilibrium with the dark matter background . in this case , the gas density and temperature profiles assume a self similar form , @xmath50 where @xmath51 has been used .
the hydrostatic equilibrium equation @xmath52 can be solved @xcite for @xmath53 for a fixed dark matter mass distribution .
@xmath54 where @xmath55 is an integration constant , @xmath56 is defined as @xmath57 and @xmath58 is the dimensionless mass within a distance @xmath59 from the center @xmath60 these integrals can be evaluated analytically @xcite for the particular cases @xmath48 and @xmath49 .
the mass - temperature normalization is given by the virial theorem @xmath61 both theoretical and numerical studies assert that the gas density profile traces the dark matter density profile in the outer regions of the halo . therefore , the slopes of these two profiles are assumed to track each other closely for @xmath62 .
this requirement fixes the polytropic index @xmath63 , and the normalization @xmath55 ( see equations 22 , 23 , and 25 in @xcite ) . the concentration parameters , @xmath44 and @xmath64 can be written in terms of the virial mass and redshift for a given cosmological model . a fitting formula based on numerical simulations @xcite is @xmath65 equation ( [ eq : c ] ) supplies a one to
one correspondence between @xmath44 and @xmath41 .
the self similar model defined above does not assign a value for the gas density normalization @xmath66 .
its value can be calculated by requiring that the ratio of the total dark matter and gas mass within some radius @xmath67 attain the universal average value of @xmath68 .
simulations without feedback from galaxy formation typically find values for the cluster gas mass fraction that are only slightly lower than the input global baryon fraction @xcite .
using cosmological parameters cited above @xmath69 where @xmath70 is the accumulated dimensionless mass of the gas analogous to @xmath71 defined in equation ( [ eq : m(x ) ] ) .
the parameter @xmath66 has only a mild @xmath67 dependence for reasonably low @xmath67 values . using a fixed @xmath72 yields a maximum error of 10% for @xmath73 $ ] .
we calculate the virial radius , @xmath42 , using the spherical top hat collapse model , i.e. we take it to be the radius of a spherical region with mean interior overdensity @xmath74 relative the critical ( obtained from the fitting formula in equation 6 of @xcite ) that encloses a total mass @xmath75 .
the corresponding angular radius of the cluster is @xmath76 where we adopt a fitting formula @xcite for @xmath77 . with this choice , the only free parameters needed to specify the full gas density profile , @xmath78 , are @xmath41 , @xmath47 , and @xmath79 .
let us now introduce a truncation of the density ( and temperature ) profile ( [ eq : y_dm ] ) to account for the shock front near the virial radius of the cluster . as pointed out above ( see eq .
[ eq : rho_gas0 ] ) , such a truncation was already needed for calculating the normalization of the gas density .
let us assume that the density has a linear cutoff between radii @xmath80 and @xmath81 . at @xmath81
the density attains the background value . since this value is small compared to @xmath82 for realistic clusters , we set both the gas density and temperature at @xmath83 to zero .
since the gas density profile traces the dark matter profile , the linear density cutoff can be imposed directly on the gas density . in what follows
, we will refer to the original density profile ( eq . [ eq : y_gas ] ) as @xmath84 .
we will denote the density profile with a cutoff by @xmath53 , defined by @xmath85 where @xmath86}/{(2d ) } & \text{if~~ } x_{\mathrm{max}}+d
< x < x_{\mathrm{max}}-d \\ 0 & \text{if~~}x_{\mathrm{max}}+d < x \\ \end{array}\right.\ ] ] for consistency , the @xmath66 gas density normalization has to be recalculated from equation ( [ eq : rho_gas0 ] ) using the mass @xmath87 enclosed by the density profiles truncated with a given choice of @xmath67 and @xmath88 ( see eq .
[ eq : m(x ) ] ) . in practice
, @xmath66 is sensitive to @xmath67 only , with the @xmath88 dependence nearly negligible .
thus , for calculating the normalization @xmath66 , we always assume @xmath89 in equation ( [ eq : rho_gas0 ] ) .
the two dimensional sze intensity profile @xmath90 , as a function of projected radial distance @xmath56 away from the cluster center , is often expressed as a small temperature change of the cmb spectrum .
the sze corresponding temperature increment @xmath91 ( which can be negative ) is defined by @xmath92 solving for the sze temperature yields @xcite @xmath93 where @xmath94 is the dimensionless frequency , @xmath95 is the compton parameter , and @xmath96 is given by @xmath97 the @xmath98 term is the relativistic correction to the frequency dependence , which is negligible at 100 ghz , but becomes important at higher frequencies ( @xmath99 ghz ) .
the compton @xmath95-parameter is defined as @xmath100 where @xmath101 is the thompson cross - section , @xmath102 is the electron number density , @xmath103 is the electron temperature , @xmath104 is the boltzmann constant , @xmath105 is the electron rest mass energy , and the integration is along the line of sight ( i.e. along @xmath106 for a given @xmath56 ) .
we calculate the electron temperature , @xmath103 , with @xmath107 using equation ( [ eq : mass - temperature ] ) , and the number density is given by @xmath108 where @xmath109 is the proton rest mass , and @xmath110 for an ionized h - he plasma with @xmath111 helium abundance by mass .
substituting ( [ eq : temperature ] ) , ( [ eq : mass - temperature ] ) , and ( [ eq : n_e ] ) in ( [ eq : compton ] ) the sz surface brightness profile ( [ eq : surfacebrightness0 ] ) is separable into a dimensionless integral with spatial dependence and a constant coefficient @xmath112 where @xmath113 and @xmath114^\gamma.\ ] ] figure [ fig : surfacebrightness ] depicts the @xmath91 profile for @xmath115 with and without a cutoff . without the cutoff , the @xmath53 and @xmath91 profile has a nonzero limit for large @xmath56 .
various choices of @xmath67 can be compared by analyzing the difference between the associated intensity profiles .
figure [ fig : surfacebrightnessdifference ] shows the difference between pairs of profiles with @xmath116 and without a cutoff , using either @xmath89 , @xmath117 , or @xmath118 for the truncated profile .
all other parameters , such as @xmath41 , @xmath47 , @xmath79 , and @xmath119 were taken to be equal .
note that the @xmath120 profile is obtained from @xmath121 by multiplying by the @xmath119 coefficient . according to equation ( [ eq : rho_gas0 ] )
, various @xmath67 choices generally lead to unequal @xmath119 values , implying a large nonzero difference between the @xmath120 profiles at @xmath122 . in principle , this difference is physical , reflecting the normalization criterion we chose ( namely that the clusters contain the average baryon fraction within the virial shock no mechanism is known to segregate baryons from dark matter on these large scales ) . since the normalization is model
dependent , one might worry that it will effect our analysis below , where we study the detectability of the virial shocks .
however , we find that the information from the normalization does not dominate our results ( see discussion below ) .
table [ tab : t ] shows the central @xmath123 values for several different cluster masses and density profile slopes , and for various frequencies .
we find that the @xmath115 dark matter profile yields @xmath124 larger @xmath123 values than the @xmath48 model .
we can elucidate the source of this increment by tracking the differences between the two models in equations ( [ eq : t(x ) ] ) and ( [ eq : t_s ] ) while fixing the other parameters at @xmath125 , @xmath126 , @xmath116 , and @xmath127 .
the difference is caused by the change in the product @xmath128 .
first , since the dark matter model with @xmath115 is concentrated more in the central region , the gas density and the gas temperature is expected to be higher in the center for @xmath115 .
indeed , @xmath129 for @xmath115 , while it is @xmath130 for @xmath48 .
second , the @xmath55 parameter is simply proportional to the central gas temperature @xmath131 ( [ eq : mass - temperature ] ) , which is somewhat higher ( by 21% ) for @xmath115 .
third , the @xmath132 factor further increases the difference by 70% . finally , the increment caused by variables localized to the center is smeared by @xmath133 , which accounts for the fact that the observation measures the projection of the intensity along the line of sight . in particular , @xmath134 for @xmath115 and @xmath135 for @xmath48 .
therefore , the resulting increase in the central sze temperature is 60% for @xmath115 .
cccccc @xmath136 & 1 & -0.019 & -0.012 & -0.003 & 0.006 + @xmath136 & 1.5 & -0.029 & -0.018 & -0.005 & 0.010 + @xmath137 & 1 & -0.106 & -0.067 & -0.018 & 0.036 + @xmath137 & 1.5 & -0.173 & -0.110 & -0.030 & 0.059 + @xmath138 & 1 & -0.609 & -0.385 & -0.106 & 0.207 + @xmath138 & 1.5 & -1.017 & -0.643 & -0.178 & 0.345 +
we shall now summarize the primary observational difficulties for detecting virial shocks . and [ sec : conclusions ] below . ]
the possible sources of contamination are detector noise , atmospheric fluctuations , and other uncertainties of astrophysical origin .
inevitable astrophysical contaminants include radio point sources , bright foreground ( and background ) clusters along the line of sight , and the primary and secondary cmb anisotropies . as pointed out in
[ sec : alma ] above , the atmospheric and detector uncertainties will not be crucial at moderate frequencies , as the alma system can reduce its noise level below @xmath1 for a @xmath139 beam at @xmath140 in an integration time of @xmath141 .
similarly , bright extragalactic foregrounds will not pose a great barrier against the observation of virial shocks .
the radio point sources cover only a negligible portion of the set of relevant pixels , and their removal reduces the efficiency of a high resolution detection by a negligible amount . the flux of the brightest clusters identified in separate wide field x - ray and sz surveys can also be subtracted from the image .
moreover , uncertainties from the imprecision of this information can be minimized by choosing galaxy clusters with no obvious overlapping foreground clusters .
the magnitude of uncertainties from primary and secondary cmb anisotropies can be assessed by the analysis of the corresponding angular power spectra .
the primary cmb anisotropy supplies the dominant power on large - scales , but is exponentially suppressed well below @xmath1 for scales under @xmath142 .
the angular radius of the clusters decreases below this scale at @xmath143 for @xmath125 and @xmath144 for @xmath145 ( eq . [ eq : theta ] ) .
we shall restrict our calculations to larger redshifts , where secondary cmb anisotropies will likely dominate .
the two largest sources of the secondary cmb anisotropies are the fluctuations from unresolved background clusters ( i.e. overlapping faint clusters of varying temperatures , sizes , and numbers along the line of sight ) and the ostriker - vishniac ( ov ) effect .
the ov contribution can be approximated by a flat band power contribution of @xmath146 @xcite . according to both theoretical and numerical studies @xcite , the rms power of the unresolved thermal sze contamination is between @xmath147 and @xmath148 on arcminute scales . for low
mass clusters this noise background is strongly non - gaussian , and there is a significant amount of uncertainty in the small - scale power spectrum , due to non - gravitational effects such as gas cooling and feedback from star formation . for a conservative estimate , we shall assume that the fluctuations around the sz profile are given by white gaussian noise with amplitude @xmath149 .
in this section , we consider the detectability of virial shocks in galaxy clusters .
the basic input to this analysis is a ( mock ) sz surface brightness map of the cluster .
we compute the sz profile @xmath150 according to equation ( [ eq : t(x ) ] ) in a mock `` fiducial '' model , in which the cluster is assumed to be spherically symmetric with a well defined edge , at the position of the virial shock .
this fiducial model is uniquely described by five parameters , specifying the global properties of the cluster @xmath151 and the location and sharpness of the virial shock @xmath152 . for concreteness , in the numerical calculations
, we use a single , fixed measurement frequency of @xmath153 .
this value was chosen to minimize the integration time needed for alma , while keeping the total sz decrement around its maximum in the nonrelativistic regime .
our basic task is to evaluate the significance , given the @xmath35k noise , at which @xmath154 can be distinguished from a different model profile @xmath155 , computed in a test model that does not include a shock .
this task immediately raises the question : what should the no edge benchmark model be for this comparison ?
the simplest approach would be to choose the underlying density profile to be identical to that in the fiducial model within @xmath156 . in the absence of a virial shock , however , the smooth density profile would extend to some large radius .
this could be expressed by choosing the same parameters @xmath157 as in the fiducial model , but replacing the location of the virial shock by @xmath158 .
however , a complication is that our procedure to self
consistently normalize the density profile ( eq.[eq : rho_gas0 ] ) would then fail , due to the divergence of the mass integral ( eq .
[ eq : m(x ) ] ) . to resolve this ambiguity
, we shall adopt either of the following two approaches to define a smooth , no edge `` test '' model ( hereafter referred to as model i and model ii , respectively ) : 1 .
the central gas density @xmath159 is chosen to equal the central density in the fiducial model .
equivalently , the central decrements @xmath119 are assumed to be the same for the two hypotheses .
the value of @xmath160 is then irrelevant , as long as it is chosen to be sufficiently large , @xmath161 ( in practice , we find that @xmath162 suffices ) .
the location of the virial shock @xmath160 is chosen to equal the radius at which the cluster gas density equals the background baryon density .
this value is between @xmath163 for typical clusters with virial mass @xmath164 , at any redshift and @xmath47 ( either 1 or 1.5 ) .
the gas density normalization is calculated self consistently , using equation ( [ eq : rho_gas0 ] ) . the profile outside @xmath160 is truncated with @xmath89 .
note that the temperature decrement scale @xmath119 and the central temperature decrement @xmath123 of this model are both different from the fiducial model ( by @xmath165 for @xmath166 for @xmath167 ) .
model ii has the advantage of a closed and self consistent theoretical description , where @xmath66 and thus @xmath168 is derived directly from a set of underlying assumptions .
however , we note that the sz brightness profile in this model differs from the fiducial model because of the `` renormalized '' value of the temperature decrement @xmath119 .
hence , the distinguishability of test model ii from the fiducial model will have an additional indirect dependence on the presence of the virial shock , through the effect of the virial shock on the overall normalization of the density profile .
this inference will then be model
dependent ( i.e. on our particular assumption about the relationship between the location of the virial shock and the normalization of the gas density ) , and also subject to uncertainties due to physical effects ( i.e. preheating ; @xcite ) that can significantly modify the inner density profile .
nevertheless , we find in our calculations below that the change in the temperature due to its renormalization is relatively small , and changes in the sz brightness profile are only significant in the inner regions .
overall , the temperature change adds nearly negligibly to the detectability of the virial shock .
hence , in practice , model ii is a useful alternative to model i. the latter , by construction , isolates the effect on the surface brightness profile of a sharp discontinuity itself .
the expected sz profile of a cluster with or without a cutoff can be calculated from equation ( [ eq : t(x ) ] ) in the fiducial model , and compared to the profile in either test model . in the next two subsections , we quantify the significance for detecting the difference between these pairs of models . in these estimates
, we will assume that the interferometer has fully synthesized an aperture , yielding a narrow ( @xmath169 ) effective point response function ( prf ) .
as discussed above , this assumption is realistic for @xmath3hr interferometric observations with alma .
thus , we will employ the simple , single
dish technique of estimating @xmath170 , in which the image of the extended source can be thought of as being composed of @xmath171 independent pixels ( where each pixel covers a solid angle of @xmath172 square arcseconds ) .
finally , as stated above , in the simplest estimates , we fix the parameters @xmath173 in model i , and the parameters @xmath151 in model ii , to be the same as in the fiducial model .
while @xmath41 and @xmath79 can be independently estimated from gravitational lensing , x ray , and optical data , the decrement normalization @xmath119 of the fiducial model will not be known a priori .
therefore , to obtain a more realistic estimate for the detectability of the virial shock , in section [ subsec : varyparameters ] below , we will relax these assumptions .
more specifically , we will obtain fits to the mock sz profile of the fiducial model , by allowing @xmath119 ( as well as the concentration parameter @xmath44 ; see further discussion below ) to be free parameters in test model i. we shall first derive a simple estimate for the significance for choosing between profiles with and without a cutoff , using the comparison between test model i and the fiducial model .
since all of the parameters of the test model are fixed , this exercise will yield an estimate for the significance of the change in the sz surface brightness maps caused by the presence of a shock front .
the presence of the shock , in general , is more difficult to infer when the test model s parameters are not assumed to be known _ a priori_. this question will be taken up in
[ subsec : varyparameters ] below .
the cutoff radii in the fiducial and test models are taken here to be @xmath167 and @xmath174 , respectively ( although our results below do not depend on the latter choice ) . the effective `` signal to noise ratio ''
is calculated by comparing the difference in the `` signal '' @xmath175 to the noise power in the regular two - dimensional angular space . the effective signal to noise ratio corresponding to a single pixel at the projected radius @xmath56 is @xmath176 in the last two lines ,
we have dropped the superscript from @xmath166 and @xmath177 because these have the same values in the fiducial and test models ( although @xmath123 and @xmath177 do depend on @xmath67 , the difference between the fiducial and test models is negligibly small ) .
we have also adopted @xmath178 for the fiducial model , in which case the maximum of @xmath179 occurs at @xmath180 ( see fig .
[ fig : surfacebrightnessdifference ] ) . for a massive cluster with @xmath125 , @xmath115 and @xmath126 , we find @xmath181 , and @xmath182 implying that @xmath183 for the single best pixel on the cluster surface , i.e. at the pixel corresponding to the edge of the cluster at the virial shock .
these values increase somewhat with increasing @xmath79 . at an effective angular resolution of @xmath184 ,
the total number of resolved pixels covered by the cluster is given by the ( solid angle extended by the circular disk within the shock radius ) divided by the ( solid angle extended by a single pixel ) : @xmath185 note that for this estimate , we consider the cluster to extend out to @xmath180 ; this will result in an underestimate , since the difference between the fiducial and test sz profiles extends out to larger radii ( see figure [ fig : surfacebrightnessdifference ] ) . since @xmath186 for @xmath126 for a massive cluster , we expect @xmath187 pixels for the angular resolution @xmath188 of alma .
of these @xmath189 pixels , @xmath190 lie along the 1pixel wide circular annulus along the virial shock .
a quick estimate for the total s / n distinguishing the two models is given by adding the maximal single
pixel s / n in quadrature , i.e. @xmath191 for a massive cluster @xmath125 at @xmath126 , we find @xmath192 , and the detection of the virial shock at @xmath193 with alma has a signal to noise ratio @xmath194 . since this value is from a single pixel
wide ring around the virial shock ,
it is strictly only a lower limit .
figure [ fig : surfacebrightnessdifference ] shows that there is an annulus of finite @xmath195 width around the virial shock with a significant @xmath196 .
@xmath195 can be estimated with the radial distance where @xmath197 drops to half of its maximum value .
a simple formula found from approximating @xmath91 to second order in @xmath56 is @xmath198 , leading to @xmath199 . for a massive cluster ,
we obtain @xmath200 and @xmath201 .
we find little dependence of this estimate on @xmath47 for choices of @xmath48 vs. @xmath115 . on the other hand
, we find a redshift dependence : larger @xmath79 decreases @xmath202 but increases @xmath203 .
the combined result is a minimum in the total @xmath170 at @xmath204 .
mass clusters have a smaller angular size and a lower sz decrement , so @xmath170 decreases with @xmath41 .
overall , the virial shock is most visible for clusters that are either close by ( @xmath205 ) , or at high redshift ( @xmath206 ) , and that have a large @xmath41 and @xmath47 . to improve the simple estimates above
, we can average the @xmath170 over the face of the cluster .
this leads to the following approximation for the average single
pixel @xmath170 : @xmath207 in appendix a , we give a rigorous derivation of the @xmath170 for distinguishing between models in the presence of a gaussian noise .
we find that equation ( [ eq : s / n(k=1 ) average ] ) , with the average taken in the interval @xmath208 , multiplied by the square root of the number of pixels , @xmath209 ( eq . [ eq : pixelno ] ) , indeed gives the formally correct answer .
accordingly , the significance of distinguishing test model i or model ii from the fiducial model is given by @xmath210 ^ 2 } & \text{(i ) } \\
\sqrt{\int{\mathrm{d}}^2x\ , [ \delta t_{s}^{\rm test } y^{\rm test}(x)-\delta t_{s}^{\rm fid}y^{\rm fid}(x)]^2 } & \text{(ii ) } \\ \end{array}\right.\ ] ] here @xmath211 is the number of pixels , and @xmath212 is the area covering the cluster s surface .
note that the only difference between models i and ii is that in the latter case , the temperature scale @xmath213 is recalculated using the cutoff location @xmath160 .
the right hand side of equation ( [ eq : s / n optimal ] ) can be compared with figure [ fig : surfacebrightnessdifference ] , which plots the integrand for model i. we evaluated the significance given by equation ( [ eq : s / n optimal ] ) for various choices of fiducial models .
recall that the fiducial model is specified by five parameters @xmath214 .
the results are listed in table [ tab : s / n optimal1 ] for three different masses ( @xmath215 ) and two different inner slopes ( @xmath216 ) .
the other three parameters in this table are fixed at the values of @xmath217 .
figure [ fig : s / n(xmax , d , z ) ] shows how the @xmath170 varies with @xmath79 and @xmath67 for these masses and slopes ( we fixed @xmath218 in this figure ) . in order to study the sensitivity of the s / n to the choice of the fiducial parameters of the virial shock itself , in table [ tab : s / n optimal2 ] we repeat each calculation in table [ tab :
s / n optimal1 ] , except we replace @xmath219 by @xmath220 .
the latter choice should also be regarded as more realistic , given that the strong virial shocks are found in simulations to be located at @xmath9 twice the virial radius ( e.g. figure 16 in bryan & norman 1998 ) , and may not be perfectly sharp . as table [ tab :
s / n optimal2 ] shows , the larger value of @xmath180 decreases the s / n by about a factor of two ( see figure [ fig : s / n(xmax , d , z ) ] ) . the main conclusion that can be drawn from table [ tab : s / n optimal1 ] and figure [ fig : s / n(xmax , d , z ) ] is that the @xmath170 is a strong function of the size of the cluster : the virial shock of a massive cluster with @xmath221 is detectable at high significance ( @xmath222 ) , but the detection becomes marginal for low
mass clusters or groups with @xmath223 , and clearly impossible for even smaller objects .
ccccccccc @xmath136 & 1 & 1.8 & 7.9 & 0.02 & 0.02 & 4.6 & 0.6 & 0.7 + @xmath136 & 1.5 & 1.8 & 4.6 & 0.2 & 0.2 & 4.7 & 0.6 & 0.8 + @xmath137 & 1 & 3.8 & 5 & 0.1 & 0.1 & 5.2 & 18 & 17 + @xmath137 & 1.5 & 3.8 & 2.9 & 1.3 & 1.1 & 5.4 & 19 & 19 + @xmath138 & 1 & 8.1 & 3.1 & 0.9 & 0.8 & 5.8 & 550 & 440 + @xmath138 & 1.5 & 8.1 & 1.8 & 8.3 & 6.9 & 6 & 570 & 480 + ccccccccc @xmath136 & 1 & 1.8 & 7.9 & 0.02 & 0.02 & 4.5 & 0.3 & 0.4 + @xmath136 & 1.5 & 1.8 & 4.6 & 0.2 & 0.2 & 4.6 & 0.3 & 0.4 + @xmath137 & 1 & 3.8 & 5 & 0.1 & 0.1 & 5.1 & 8.9 & 10 + @xmath137 & 1.5 & 3.8 & 2.9 & 1.2 & 1.1 & 5.3 & 9.6 & 12 + @xmath138 & 1 & 8.1 & 3.1 & 0.9 & 0.8 & 5.7 & 280 & 240 + @xmath138 & 1.5 & 8.1 & 1.8 & 7.9 & 7.0 & 5.9 & 300 & 280 + + we also find that the @xmath170 is nearly identical for different choices of @xmath47 and @xmath88 .
however , figure [ fig : s / n(xmax , d , z ) ] shows that @xmath170 decreases with @xmath67 , and it is a non monotonic function of redshift with a minimum at @xmath224 .
two of our findings are somewhat counterintuitive : cluster edges are equally visible for smeared edges , and are more visible for more distant clusters , even though these have a smaller angular size .
the first of these results can be explained by recalling the definition of the cutoff .
increasing @xmath88 does not change the total cluster sz intensity .
if the cutoff shape is known exactly prior to the observation , as it is assumed in this section , increasing @xmath88 ( while fixing @xmath119 ) does not come closer on average to a profile without a cutoff . to explain the second peculiarity , the behavior of @xmath170 as a function of @xmath79 , we considered the @xmath79dependence of each individual factor that determines the @xmath170 . the most puzzling feature , i.e. the increase of @xmath170 towards high redshifts at @xmath225 , is attributable to an increase in the temperature scale and in the central temperature decrement @xmath123 ( shown in figure [ fig : z ] ) .
although the angular radius @xmath226 decreases somewhat towards high@xmath79 , the temperature increase dominates . as a result ,
more distant clusters at @xmath225 have a higher contrast , which makes the virial shock more visible even though the clusters have a smaller angular size .
will have a cutoff .
however , at the @xmath188 resolution of alma , this critical redshift is well beyond the epoch when clusters form in a sensible cosmology .
redshift ( @xmath227 ) , the situation is reversed , and the strong increase towards @xmath228 in the angular size is the dominant effect , making the virial shock of very nearby clusters more detectable .
table [ tab : s / n optimal1 ] also compares the @xmath170 results for model i and ii .
the two models lead to approximately equal @xmath170 detection ratios .
this is somewhat surprising , since the renormalization in the temperature scale @xmath119 and the difference in the profile shape @xmath121 in model ii both introduce additional differences from the surface brightness @xmath91 in the fiducial model .
we find that these additional differences can be significant , as high as @xmath229 for each pixel near the central region @xmath230 .
however , the difference is only at the level of @xmath231 per pixel at the radius @xmath232 that dominates the cumulative @xmath170 .
+ in conclusion , the estimates in this section indicate that the temperature decrement difference across a virial shock front causes a difference of approximately @xmath233k in the sz surface brightness maps .
this is a small difference , approximately at the level of the expected noise .
hence , in single neighboring pixels , the shock fronts will not stand out to be detectable .
however , shock fronts are coherent structures extending over many pixels ; in the simple spherical models considered here , the cumulative @xmath170 ratio is sufficiently high to infer the presence of the shock front with a high resolution instrument , such as alma .
the presence of shocks in medium
sized to massive clusters should be clearly inferable , as long as the shocks are located near the virial radius .
if the density cutoff is farther in the outer region , the virial shock will only be observed for relatively massive clusters , with @xmath234 .
the detection significance is nearly independent of @xmath47 and @xmath88 .
the virial shocks of clusters at @xmath224 are least detectable , with nearly an order of magnitude increase in the @xmath170 for @xmath235 or @xmath236 compared to the minimum value at @xmath126 assuming that the noise is uncorrelated and gaussian for different frequencies , measuring at many different frequencies the signal to noise ratio can be increased by the square root of the number of available independent frequency channels .
the main deficiency of the signal to noise ratio analysis in the previous subsection is that at least some of the parameters of the test model will not be known a priori .
this , in general , will make the cluster s virial shock more difficult to detect , since one may be able to adjust the variable parameters of the no edge test model to better mimic the fiducial model that includes the shock .
the mass @xmath41 and redshift @xmath79 of the cluster can be independently estimated ( e.g. from gravitational lensing , x ray , and optical data ) , and the inner density profile slope @xmath47 has a relatively small impact on our conclusions .
hence , we still fix the values of these parameters in this section .
although the sz temperature decrement scale @xmath237 can , in principle , be derived in the fiducial model , it will not be known a priori . we therefore allow it to be a free parameter .
in addition , the concentration parameter of the cluster @xmath44 , which , in principle , can be ultimately derived from simulations , is still somewhat uncertain .
therefore , we allow @xmath238 to be the second of our two free parameters of the test model in this section .
the maximum radius used for the comparison of the sz profiles including and lacking a cutoff is @xmath239 , the value used in model ii previously .
recall that this is the radius at which the baryon fraction equals the global value @xmath240 .
a different complication is that our analysis relies on mock data , rather than actual data .
therefore , in the presence of noise , our `` data '' should not be a fixed set of numbers ( i.e. the profile @xmath91 in the fiducial model ) .
rather , it must be a ( set of ) probabilistic variables that have a distribution and finite width .
a rigorous signal to noise analysis that incorporates these two complications is presented in appendix b. in this case , the @xmath170 of distinguishing the test model from the fiducial model is itself a random variable ( rather than a fixed number , as we have hitherto assumed ) .
however , in appendix b we show that the expectation value of @xmath170 obeys the following bounds : @xmath241 where @xmath242 is defined by @xmath243 ^ 2.\ ] ] the term @xmath244 refers to the two variable parameters of the test model ( i.e. @xmath237 and @xmath238 ) , and @xmath245 refers to the fiducial values @xmath246 , @xmath247 , @xmath248 , and @xmath249 .
this distinction is necessary since the corresponding parameters need not be equal .
for a conservative estimate we shall use the lower bound in equation ( [ eq : s / n bounds ] ) .
the expected value of the signal to noise ratio is therefore estimated by @xmath250 ^ 2}.\ ] ] this should be compared with equation ( [ eq : s / n optimal ] ) describing the signal to noise ratio for fixed parameter values .
the differences are the overall @xmath251 factor and taking the minimum over the @xmath244 parameter space .
we shall denote the best fitting test parameters , i.e. at which the integral is minimal , by @xmath252 .
equation ( [ eq : signal to noise advanced ] ) measures the probability that the virial shock of the cluster is observable even in the face of two additionally variable parameters .
+ the numerical results presented in the last section in table [ tab : s / n optimal1 ] and figure [ fig : s / n(xmax , d , z ) ] considered @xmath244 parameter values that were equal to the corresponding @xmath245 parameter values . here
we consider the fiducial model with @xmath253 and @xmath254 or @xmath255 .
the fiducial values @xmath256 are calculated from equations ( [ eq : c ] ) and ( [ eq : t_s ] ) as before .
the parameters @xmath244 are assumed to be the same for @xmath151 as in the fiducial model , but , in contrast with the previous section , @xmath257 are allowed to vary relative to the fiducial values .
the fractional deviation of the signal to noise ratio as compared to the original value of test model i for various @xmath244 parameters is plotted in figure [ fig : h2_minimum34 ] . the @xmath251 factor decrease because of the increase in the average noise power is not included in the contour levels for clarity .
thus , the contours have the value of 1.0 at the fiducial parameters .
the best fitting parameters @xmath252 correspond to the values at the minimum of s / n .
the most important result from the analysis displayed in figure [ fig : h2_minimum34 ] is that the signal to noise does not decrease significantly , even in the best
fitting model , relative to the values listed in the previous section ( table [ tab : s / n optimal1 ] and figure [ fig : s / n(xmax , d , z ) ] ) . by obtaining the best fit for various @xmath67
we find that the maximum decrease is a factor of @xmath258 , and @xmath259 for @xmath260 , and @xmath255 respectively .
the virial shock detection probabilities therefore remain significant for @xmath261 , but become marginal ( @xmath262 ) for low
mass clusters ( @xmath263 ) .
another interesting feature shown in figure [ fig : h2_minimum34 ] is that the best fitting parameters can deviate significantly from the fiducial values .
changing @xmath41 , @xmath67 and @xmath88 shows that the @xmath170 of particular clusters can have saddle points and multiple local minima in terms of the @xmath244 parameters .
we find that this is the case , for example , for @xmath264 and @xmath265 , where the local minimum near the fiducial values is not global , and the global minimum @xmath266 instead lies at a much larger sz temperature and a much smaller concentration .
this behavior follows from the fact that decreasing @xmath44 , while fixing all other parameters , produces a steeper @xmath267 profile on log - log scale in the outer regions , a better approximation of the fiducial model with a virial shock . on the other hand ,
the decrease in @xmath44 also slightly decreases the sz temperature in the inner regions , which can be compensated by increasing @xmath119 .
overall , a test model profile with a smaller @xmath268 and larger @xmath44 comes closer to the fiducial model . in appendix c
, we present a simple method to obtain analytical estimates for the best fitting @xmath252 parameters of a false test model .
this approach is fully general , and gives a numerically efficient way to compare two arbitrary models with arbitrary numbers of parameters .
our method is somewhat analogous to a fisher matrix approach .
the standard fisher matrix treatment serves to estimate the expected uncertainty of the parameters of a correct model .
in contrast , the method presented in appendix c gives an estimate of the best fitting parameters in an incorrect model ( and does not address the uncertainty of these best fitting values ) .
in the previous section , we have examined the signal to noise ratio for distinguishing test models from the fiducial model .
the primary aim was to make a distinction between the two basic possibilities , i.e. whether the cluster does or does not have an edge .
a more ambitious problem is to examine the precision with which the location and sharpness of the edge can be measured , i.e. the precision of determining the parameters that yield the profile @xmath154 .
we calculate the uncertainty of the parameter estimation using both a nonlinear likelihood function directly , and with an approximate fisher matrix method .
we adopt the fiducial model of the previous section to predict the sz temperature decrement profile . as before
, the fiducial model includes an edge , and is specified by the five parameters @xmath41 , @xmath79 , @xmath47 , @xmath67 , and @xmath88 whose fiducial values will here be collectively denoted by @xmath245 .
we then compare the temperature decrement in the fiducial model , @xmath269 , to two different sets of test models .
the test models in this case are identical to the fiducial model ( i.e. , they include an edge ) , except we allow a subset of the parameters to have values @xmath270 .
these two cases are analogous to the test models i and ii in the previous sections .
in the first case ( model i ) , we consider a set of four unknowns @xmath119 , @xmath44 , @xmath67 , and @xmath88 . in the second case ( model ii )
, we consider only two free variables @xmath67 and @xmath88 , and fix @xmath119 and @xmath44 at the values calculated in the fiducial model using equations ( [ eq : t_s ] ) and ( [ eq : c ] ) .
the other parameters describing the cluster , @xmath41 , @xmath79 , and @xmath47 , are held fixed in both cases .
now let us suppose that a temperature decrement @xmath271 is measured , and with no prior information , a parameter @xmath244 is to be chosen that best describes the data .
due to random noise , the measurement yields a parameter estimator with some uncertainty .
the parameter estimator and its uncertainty can be obtained with the maximum likelihood test , following appendix a. the likelihood function is @xmath272 ^ 2\right ) .
\nonumber\end{aligned}\ ] ] the constant normalization pre - factor is irrelevant for the likelihood ratio test , which we hereafter omit from the likelihood function . the log likelihood is @xmath273 ^ 2 .\ ] ] where @xmath171 and @xmath274 are the total number of pixels and the surface area of the cluster .
the parameters are then chosen to maximize @xmath275 , or equivalently by applying a least squares fit to the data .
the signal to noise ratio is obtained by the analysis of the projection of the sz observation profile on the subspace spanned by the fiducial model profiles with arbitrary parameters .
the noise power for @xmath276 variable parameters is the variance on this subspace , @xmath277 , and the signal to noise ratio is therefore @xmath278 ^ 2\ ] ] equation ( [ eq : false signal to noise ] ) measures the extent that the parameter estimator is @xmath244 instead of @xmath279 , the true ( and maximum likelihood ) value .
@xmath280 follows @xmath281-statistics with @xmath276 degrees of freedom , leading to @xmath282 for @xmath283 errors .
given a true parameter set @xmath245 , the region within @xmath284 confidence ( 95% ) for example is the set of @xmath244 values for which @xmath285 .
the uncertainty of the parameter estimation can therefore be read off directly from equation ( [ eq : false signal to noise ] ) .
figure [ fig : parameter variance12 ] shows contour diagrams at fixed values of this uncertainty in model ii , assuming the fiducial parameters @xmath253 with @xmath286 ( upper panel ) or @xmath287 ( lower panel ) .
the contour plots allow @xmath67 and @xmath88 to vary , while all other parameters are held fixed at the fiducial values .
+ figure [ fig : parameter variance12 ] depicts the likelihood contours for model ii with the direct evaluation of the @xmath288 .
an estimate for these contours is provided by the fisher information matrix , @xmath289 assuming that the likelihood distribution is gaussian near the peak likelihood , we can use confidence limits for gaussian statistics ( i.e. , @xmath281 ) to obtain @xmath290 and @xmath291 confidence regions .
the fisher matrix greatly speeds up the computations , and allows us to consider additional free parameters ( model i ) .
the minimum expected variance is related to the fisher matrix by the cramer - rao bound @xcite @xmath292 where equality holds if the distribution is well approximated by a gaussian distribution .
tables [ tab : parameter variance2 ] and [ tab : parameter variance1 ] show the 68% and 95% significance uncertainties for models i and ii , as calculated from the fisher matrix .
generally , increasing the number of free variables can increase the uncertainty of the original variables , since a model with additional free parameters can better mimic the fiducial model for a given mock observation .
however , our calculations indicate that this is not the case in our case : the uncertainties on @xmath67 and @xmath88 are nearly the same in models i and ii . this is perhaps unsurprising , since one would not expect that a change in @xmath44 or @xmath119 ( the additional free parameters in model i ) can mimic changes in either @xmath67 and @xmath88 .
indeed , the fisher matrix is decoupled in the corresponding subspaces , as the @xmath293 , @xmath294 , @xmath295 , @xmath296 components are negligible compared to the diagonal elements .
the uncertainties in the parameter estimators are @xmath297 for @xmath298 .
therefore @xmath44 , @xmath119 , and @xmath67 will be precisely obtained with alma for the majority of the clusters , while similar precision for @xmath88 is possible for only massive clusters .
we note that in principle , it is possible to deduce the value of @xmath47 directly from the @xmath299 relation .
figure [ tab : parameter variance2 ] provides confidence in our conclusions from the fisher matrix method : the @xmath283 contours in @xmath300 are well - approximated by ellipses , if the fiducial values obey @xmath301 , e.g. @xmath302 , @xmath303 , and @xmath304 .
therefore , in these cases , the inequality ( [ eq : cramer - rao ] ) assumes equality for the parameter uncertainty .
we find that the parameter distribution around @xmath305 is distorted , and higher confidence level contours for arbitrary @xmath88 are banana shaped in the @xmath80 plane .
cccccccc @xmath136 & 1 & 1.8 & 7.9 & 23 & 2 & 0.5 & - + - & - & - & - & - & 6.5 & 17 & 68 + - & - & - & - & - & 13 & 34 & 95 + @xmath136 & 1.5 & 1.7 & 4.6 & 170 & 2 & 0.5 & - + - & - & - & - & - & 6.1 & 16 & 68 + - & - & - & - & - & 12 & 32 & 95 + @xmath137 & 1 & 3.8 & 5.0 & 140 & 2 & 0.5 & - + - & - & - & - & - & 0.20 & 0.55 & 68 + - & - & - & - & - & 0.41 & 1.1 & 95 + @xmath137 & 1.5 & 3.8 & 2.9 & 1200 & 2 & 0.5 & - + - & - & - & - & - & 0.19 & 0.52 & 68 + - & - & - & - & - & 0.38 & 1.04 & 95 + @xmath138 & 1 & 8.1 & 3.1 & 860 & 2 & 0.5 & - + - & - & - & - & - & 0.006 & 0.018 & 68 + - & - & - & - & - & 0.013 & 0.036 & 95 + @xmath138 & 1.5 & 8.1 & 1.8 & 7900 & 2 & 0.5 & - + - & - & - & - & - & 0.006 & 0.017 & 68 + - & - & - & - & - & 0.012 & 0.035 & 95 + cccccccc @xmath136 & 1 & 1.8 & 7.9 & 23 & 2 & 0.5 & - + - & - & - & 1.0 & 15 & 6.5 & 17 & 68 + - & - & - & 2.1 & 29 & 13 & 34 & 95 + @xmath136 & 1.5 & 1.7 & 4.6 & 170 & 2 & 0.5 & - + - & - & - & 0.83 & 132 & 6.1 & 16 & 68 + - & - & - & 1.7 & 263 & 12 & 32 & 95 + @xmath137 & 1 & 3.8 & 5.0 & 140 & 2 & 0.5 & - + - & - & - & 0.15 & 6.4 & 0.21 & 0.55 & 68 + - & - & - & 0.30 & 13 & 0.41 & 1.09 & 95 + @xmath137 & 1.5 & 3.8 & 2.9 & 1200 & 2 & 0.5 & - + - & - & - & 0.11 & 67 & 0.19 & 0.52 & 68 + - & - & - & 0.22 & 135 & 0.39 & 1.05 & 95 + @xmath138 & 1 & 8.1 & 3.1 & 860 & 2 & 0.5 & - + - & - & - & 0.013 & 2.2 & 0.007 & 0.018 & 68 + - & - & - & 0.026 & 4.4 & 0.013 & 0.036 & 95 + @xmath138 & 1.5 & 8.1 & 1.8 & 7900 & 2 & 0.5 & - + - & - & - & 0.011 & 27 & 0.006 & 0.017 & 68 + - & - & - & 0.023 & 55 & 0.013 & 0.035 & 95 + our calculations show that increasing @xmath67 increases the uncertainties @xmath306 and @xmath307 , while increasing @xmath88 only slightly increases @xmath306 and leaves @xmath307 unchanged .
it is clear that the choice of @xmath47 does not alter the @xmath306 and @xmath307 uncertainties .
changing @xmath79 yields a maximum in @xmath308 and @xmath307 around @xmath309 .
we find that the parameter uncertainties decrease by a factor of @xmath310 at @xmath235 or @xmath236 , compared to the @xmath309 value .
we have shown that the forthcoming atacama large millimeter array ( alma ) is well suited for studies of the intra
cluster medium density distribution using the sunyaev - zeldovich ( sz ) effect .
the angular beam diameter and sensitivities are predicted to reach @xmath139 and @xmath1 for this system , which exceed present detector resolutions by more than two orders of magnitude .
the sz decrement profile of a rich galaxy cluster , observed at 100 ghz for @xmath930 hours , will have a sufficiently high signal - to - noise ratio to make inferences about the gas distributions ( and , by inference , about the dark matter distributions ) .
we solved the equations of hydrostatic equilibrium for the self
similar gas density in the dark matter background .
the dark matter profile was parameterized by its inner slope @xmath47 , and every calculation was evaluated for the two values common in the literature , i.e. @xmath48 and 1.5 . within this framework ,
the density as a function of radial distance is small but nonzero , even in the outer regions . as a modification
, we introduced a linear cutoff in the density profile , where the density falls to zero within a finite radial distance .
this assumption mimics the presence of strong virial shocks , predicted in cosmological structure formation theories ( see @xcite and @xcite ) .
we have calculated the sz decrement profiles of clusters with and without the presence of a sharp density cutoff near the virial radius .
we assumed @xmath1 flat - power gaussian noise .
we set up the likelihood function and defined the decision rule to see whether our fiducial model with a virial shock can be distinguished from a test model without a virial shock .
it is important to emphasize that this test is practically model independent : it is only weakly sensitive to the central regions , where cluster models tend to differ , with most of the sensitivity arising from the outer regions , where the various cluster models in the literature are very similar .
we calculated the signal to noise ratio in two ways , with either four ( model i ) or two ( model ii ) free parameters .
these parameters are the radial distance to the virial shock , @xmath67 , the radial thickness of the cutoff , @xmath88 , the concentration parameter , @xmath44 , and the temperature decrement scale , @xmath119 , for model i. for model ii , @xmath44 and @xmath119 were calculated from theory assuming no scatter around the correct values .
other parameters , such as the virial mass , @xmath41 , the redshift , @xmath79 , and the logarithmic slope of the inner dark matter profile , @xmath47 , can , in principle , be obtained from independent measurements , and were not varied .
the signal to noise ratios agree to within 20% in model i and model ii .
the overall value is s / n=470 for a rich cluster of virial mass @xmath311 and we find s / n=16 for a lower mass cluster with @xmath145 .
we also considered a more realistic case , in which the parameters of the cluster with and without a virial shock were fitted independently for a given measurement .
we find a large systematic bias in the parameters @xmath44 and @xmath119 for the model without a virial shock , provided that the cluster does have a virial shock within approximately two virial radius .
nevertheless , the signal to noise ratio decreases only by @xmath312 at most for detecting the virial shock .
therefore , the virial shocks remain detectable for @xmath313 clusters . finally , we examined the precision to which the parameters of the virial shock could be obtained .
we find the typical values for the precisions for low
mass ( @xmath314 ) and rich ( @xmath315 ) clusters to be @xmath316 and @xmath317 , respectively .
these results assume the cluster is at @xmath126 , and it is surrounded by a virial shock at twice the virial radius ( @xmath116 ) that has a finite width @xmath318 .
if the shock was located closer in ( @xmath193 ) and was sharper ( @xmath218 ) , these uncertainties would improve to @xmath319 and @xmath320 , respectively . since there is a one to one theoretical correspondence between @xmath44 and @xmath47 , with @xmath321 , sz measurements should help in determining the value of @xmath47 .
this should serve as a valuable diagnostic of the inner gas profile ; in particular , of the presence of any excess ( non gravitational ) entropy that could strongly modify the inner profile .
our analysis is based on a simplified , spherically symmetric model for the cluster . in these models ,
the virial shock appears in the sz surface brightness maps as a full ( 2@xmath322 azimuthal angle ) ring like feature . in reality
, the gas infall into the cluster potential should be inhomogeneous and anisotropic . as a result ,
the radial location and the strength of the virial shock will vary in different directions away from the cluster s center @xcite .
we partially accounted for this effect by allowing the virial shock to have a finite width ( @xmath88 ) , which `` smears out '' the ring like feature mimicking the effect of projecting shocks of varying radial locations and strengths .
simulations also show that gas is falling into the cluster potential along overdense filaments , and that the strongest shocks occur along these filaments .
if the projected ring like feature extended over only a fraction @xmath323 of the full @xmath324 azimuthal angle ( corresponding to the projected areas of the filaments ) , then our predicted @xmath170 for detecting the virial shock would be further reduced by a factor @xmath325 .
the virial shock will nevertheless remain detectable for a @xmath326 cluster at a significance of @xmath327 , unless @xmath328 .
our results suggest that sz decrement measurements with alma can reveal virial shocks around galaxy clusters , and can determine the location and size of these shocks with high precision .
this will provide a unique constraint on theories of large scale structure formation .
we thank gil holder , greg bryan , david helfand , and sndor molnr for useful discussions .
we acknowledge support from otka through grant nos .
t037548 , t047042 , and t047244 .
in this appendix , we give a formal summary of the treatment of the following general problem .
consider a system that obeys laws that predict a set of observables @xmath329 .
what is the significance at which a false hypothesis , predicting a different set of observables , @xmath330 , can be rejected , given the presence of gaussian random noise ?
a measurement of the observable in @xmath331 distinct directions yields a discrete sample @xmath332 where @xmath333 $ ] .
the set @xmath334 is an element of a @xmath171-dimensional vector space and will be denoted by @xmath335 .
similarly , let @xmath329 and @xmath330 denote the discrete sample of the hypotheses functions @xmath336 and @xmath337 .
in our example , @xmath335 denotes a measured sze temperature profile , whereas @xmath329 and @xmath330 denote the discrete temperature profiles predicted in models with and without a cutoff .
the vectors @xmath329 and @xmath330 depend on the parameters describing the cluster and the cutoff .
we now derive the signficance ( or `` signal to noise ratio '' ) of distinguishing between two models .
if the real signal arriving to the detector was @xmath338 , then the detector measures the data @xmath339 where @xmath340 is a random variable , corresponding to the noise . let us assume white gaussian noise of variance @xmath341 . in this case , the probability of detecting @xmath335 , given that the incoming signal is @xmath342 , is @xmath343 , \ ] ] where the arithmetic ( difference and scalar product ) of the k - dimensional vectors was used , i.e. @xmath344 two hypotheses can be distinguished using @xmath345 , which can be obtained from @xmath346 and the a priori probabilities @xmath347 by applying the bayes theorem for conditional probabilities .
most common decision rules , such as the maximum posteriori , the neyman - pearson , or the minimax decisions ( e.g. @xcite ) involve constraints on the likelihood ratio , @xmath348 substituting ( [ eq : p(y|s ) ] ) in ( [ eq : l(y)def ] ) the log likelihood becomes @xmath349 the decision is made in favor of the test hypothesis if the likelihood ratio exceeds a given threshold .
the likelihood depends on the @xmath350 component along @xmath351 .
the term @xmath351 is referred to as the matched filter , which is the optimal filter for the distinction of the two hypotheses in white gaussian noise . since the noise distribution @xmath352 is spherically symmetric in the k - dimensional vector space , it has the same power @xmath341 along any basis .
this implies that the noise power for the likelihood detection rule is the noise power for a single bin , i.e. @xmath353 . thus increasing the sample size increases the signal power , but leaves the relevant noise contribution the same .
the signal power is @xmath354 where @xmath355 and @xmath356 denotes the area enclosed by the neighboring @xmath357 points .
we have assumed that the resolution of the measurement is fine enough to approximate the sum with the integral .
therefore , the signal to noise ratio is @xmath358 the signal to noise ratio is an important measure of the significance of the test .
let us define @xmath359 as a one - dimensional random variable by @xmath360 for which @xmath361 is a standard normal distribution , whereas @xmath362 is a normal distribution with mean @xmath170 and unit variance .
thus , if the fiducial model is true , @xmath363 follows @xmath281 statistics , and the significance of the test can be obtained by comparing @xmath364 with the @xmath281 confidence levels . in our case , for a detector with angular resolution @xmath184 and a cluster of apparent angular virial radius @xmath226
the surface element is @xmath365 , where @xmath366 is the total number of pixels and @xmath212 is the surface area of the cluster .
therefore , @xmath367
^ 2}.\ ] ] is the signal to noise ratio for the decision between the two hypotheses .
here we address the following problem , which is a generalized version of the problem posed in appendix a above .
consider again a system that obeys laws that predict a set of observables @xmath368 .
what is the significance at which a false hypothesis , predicting a different set of observables , @xmath369 , can be rejected , given the presence of a gaussian random noise , _ if the parameters of the false test model are not known a priori and can be freely adjusted _ ?
assume that the original signal ( i.e. without noise ) is @xmath368 with a set of @xmath276 parameters , @xmath370 , and the false hypotheses is described by @xmath371 parameters , @xmath244 .
the measurement @xmath372 can be used to give an estimate of @xmath373 . as in appendix
a , @xmath335 is the collection of @xmath171 observables ( e.g. the sz brightness of the @xmath171 pixels ) , and @xmath340 is a @xmath171-dimensional gaussian random variable . denote the estimated parameter of @xmath368 by @xmath245 and the estimated parameter of @xmath330 by @xmath244 .
once the parameters have been obtained , @xmath374 and @xmath375 can be fixed at the corresponding values .
then the likelihood ratio for fixed parameters can be used , according to equation ( [ eq : lnl(y ) ] ) , and the decision is made in favor of the test model rather than the fiducial model exactly if the likelihood is above a given threshold .
since @xmath245 is obtained by minimizing @xmath376 , the distribution of @xmath377 from ( [ eq : y_distribution ] ) is @xmath378 where @xmath373 is the real parameter of the signal without the noise and @xmath379 is an @xmath276 dimensional gaussian random variable . is a gaussian random variable in terms of the set of parameters @xmath279 only if the various choices for these parameters span a flat ( linear ) subspace .
this is an adequate approximation , provided that the radius of curvature is much less than the r.m.s .
variance of the parameters , @xmath380 .
] @xmath381 has zero mean and variance @xmath382 .
similarly , @xmath244 is obtained by minimizing @xmath383 .
the distribution of @xmath375 is @xmath384 where @xmath252 is the @xmath244 parameter value at which @xmath385 is minimal .
@xmath386 is defined by equation ( [ eq : h2(p2 ) ] ) , which is an @xmath371-dimensional gaussian random variable , with zero mean and variance @xmath387 .
note that @xmath379 and @xmath386 are strongly correlated since both values are derived from the same measurement @xmath335 .
the signal power @xmath388 is therefore a random variable .
let us define the empirical , most probable and @xmath252 .
the term most probable refers to the parameter distribution not the signal power distribution . ] and expected signal powers by @xmath389 the expected signal power can be written in terms of @xmath390 using eqs .
( [ eq : h1(p1 ) ] ) and ( [ eq : h2(p2 ) ] ) @xmath391 ^ 2\rangle \\ & = s_0 ^ 2 + \langle(\bm{\delta}_2 - \bm{\delta}_1 ) ^2 \rangle\\ & = s_0 ^ 2 + \langle\bm{\delta}_2 ^ 2\rangle + \langle\bm{\delta}_1 ^ 2\rangle - 2 \langle \bm{\delta}_1 \bm{\delta}_2\rangle\\ & = s_0 ^ 2 + ( n^{\rm fid } + n^{\rm test})\sigma_n^2 - 2 \langle \bm{\delta}_1 \bm{\delta}_2\rangle .
\label{eq : signal exact}\end{aligned}\ ] ] in particular if @xmath392 , then @xmath393 .
this term can be obtained , for given @xmath329 and @xmath369 hypothesis functions by @xmath394 for arbitrary numbers of parameters , the expected value of @xmath395 has a more complicated algebraic form , but can always be obtained for a specific choice of @xmath329 and @xmath330 functions . for simplicity we shall restrict only to the inequality @xmath396 leading to @xmath397 the noise power for choosing between @xmath329 and @xmath330 ( with an arbitrary @xmath245 or @xmath398 ) is the variance of the measurement along @xmath399 , since the likelihood ratio ( [ eq : lnl(y ) ] ) depends on only this component . therefore @xmath400
^ 2\right>\\ & = \sigma_n^2 + \left < \left[\frac{{\bf h}^{\rm test}(p^{\rm test})-{\bf h}^{\rm fid}(p^{\rm fid})}{||{\bf h}^{\rm test}(p^{\rm test})-{\bf h}^{\rm
fid}(p^{\rm fid})|| } \cdot \bm{\delta}_1 \right]^2\right > .\end{aligned}\ ] ] for large signal to noise ratios this can be approximated to lowest order in @xmath379 .
@xmath401 ^ 2\right>\\ & = \sigma_n^2 + \langle(\cos\theta)^2\rangle \sigma_n^2 , \end{aligned}\ ] ] where @xmath402 is the angle between the vector @xmath403 and @xmath404 . in general for any @xmath405 @xmath406 equation ( [ eq : noise ] ) is a natural consequence of the fact that there are two sources of uncertainties for the hypothesis test , corresponding to the uncertain estimates of the parameters of the fiducial and test models . in principle
, these uncertainties can add up constructively to create a @xmath407 variance in the likelihood ratio , but the noise power is at least the noise of the fixed models of appendix a , @xmath408 . in particular , for the sze brightness hypotheses , the test model is a special case of the fiducial model , with the fiducial model having two additional parameters describing the cutoff at the virial shock . in this case for @xmath409 and @xmath410 , it can be shown that @xmath411 independent of @xmath370 .
therefore @xmath412 although the expected signal to noise ratio can be calculated explicitly for a given parameter choice using equation ( [ eq : signal exact ] ) and ( [ eq : noise exact ] ) , it is useful to define its the theoretical bounds independent of the given form of the hypotheses . comparing equations ( [ eq : signal ] ) and ( [ eq : noise ] ) , we obtain @xmath413 where @xmath390 is given by equation ( [ eq : s_0 ] ) .
the high bound is approached when @xmath414 tends to be parallel to @xmath415 , the low bound is approached when @xmath414 is orthogonal to @xmath415 near @xmath416 and @xmath417 .
here @xmath418 denotes the sub space of the k dimensional vector space that is spanned by the variable parameters .
here we present an estimate for the values of the best fitting parameters of a false hypothesis .
this problem can be contrasted with the application of the fisher matrix . while the latter is a method to obtain the variance on the parameters of the true model , here we seek an approximation of the expectation values for the best - fitting parameters of a false test model ( regardless of the uncertainties around this best fit , and whether the best is an acceptable fit ) .
finding @xmath252 , the minimum value in the @xmath244 parameter space in general can be computationally tedious .
our treatment is general as long as the fiducial hypotheses has all of the parameters of the test hypotheses plus some additional parameters and if the two hypotheses deviate by only a reasonably small amount . in particular , as displayed in the panels of figure [ fig : h2_minimum34 ] , the numerical value obtained naively by @xmath419 is a fair zeroth approximation for the two cluster models in most cases .
a better approximation can be obtained by expanding @xmath420 ^ 2 $ ] to second order around @xmath421 and finding its minimum in terms of @xmath244 .
@xmath422 x_jx_k \nonumber\end{aligned}\ ] ] where it is assumed that @xmath330 and @xmath423 are evaluated at @xmath421 and @xmath245 , respectively .
the indices @xmath424 and @xmath425 run across the test model parameters , and @xmath426 , furthermore @xmath427 let us denote the quadratic coefficient by @xmath428 equation ( [ eq : p expansion ] ) gives the deviation of the signal to noise ratio if the test model parameters are modified from the fiducial values .
the expected signal to noise corresponds to the minimum value in terms of @xmath357 , at which the @xmath357 derivative must vanish @xmath429 ^ 2\approx \\ & 2(\partial_k{\bf h}^{\rm test})\cdot({\bf h}^{\rm test}-{\bf h}^{\rm fid } ) + 2m_{kj}x_j=0 .\nonumber\end{aligned}\ ] ] since @xmath430 , the solution of this equation gives the next approximation of @xmath431 .
equation ( [ eq : p expansion derivative ] ) is solved by inverting the coefficient matrix .
thus @xmath432 @xmath433 is the improved approximation of the @xmath434 parameter of @xmath435 .
note that the @xmath119-dependence of the hypotheses is linear , implying @xmath436 and @xmath437 for all @xmath244 .
the @xmath238 parameter derivatives are to be calculated numerically using equation ( [ eq : t(x ) ] ) . in conclusion , finding the minimum in equation ( [ eq : signal to noise advanced ] ) is simplified to evaluating parameter derivatives of the @xmath330 function at only @xmath438 .
figure [ fig : h2_minimum34 ] shows how the best fitting @xmath439 parameters are related to the naive choice @xmath245 when detecting virialization shocks with the sz effect .
the @xmath170 function has a minimum at @xmath435 . finding the critical point with equation ( [ eq : p approximation ] ) yields the correct global minimum for @xmath440 for any @xmath300 and @xmath441 for any @xmath300 unless @xmath442 .
however , for @xmath443 there are multiple local minima and saddle points .
the @xmath444 approximation of equation ( [ eq : p approximation ] ) yields the local minima in the vicinity , which is _ not _ the global minimum . for @xmath445 and arbitrary @xmath300
, the @xmath446 value corresponds to the saddle point .
again , the @xmath447 approximation of equation ( [ eq : p approximation ] ) breaks down .
the saddle points can be identified by calculating @xmath448 @xmath449 the saddle point can be eluded if @xmath450 is calculated and a step is made towards @xmath446 only if @xmath451 .
if it is negative , then a step is made downhill along the negative gradient @xmath452 .
repeating equation ( [ eq : p approximation ] ) or the downhill steps gives a numerical efficient way of obtaining better and better approximations of the best fit test parameter values near the corresponding fiducial parameters .
further study shows that this algorithm leads to the correct @xmath435 value in most cases .
only the fiducial parameters around @xmath453 , with @xmath454 or @xmath455 have non - global minimum in the close neighborhood of @xmath373 . in this case
the best fitting test model parameters have to be obtained by evaluating equation ( [ eq : s_0 ] ) at multiple parameter values , scanning the parameter space . | in cosmological structure formation models , massive non linear objects in the process of formation , such as galaxy clusters , are surrounded by large - scale shocks at or around the expected virial radius .
direct observational evidence for such virial shocks is currently lacking , but we show here that their presence can be inferred from future , high resolution , high sensitivity observations of the sunyaev - zeldovich ( sz ) effect in galaxy clusters .
we study the detectability of virial shocks in mock sz maps , using simple models of cluster structure ( gas density and temperature distributions ) and noise ( background and foreground galaxy clusters projected along the line of sight , as well as the cosmic microwave background anisotropies ) .
we find that at an angular resolution of @xmath0 and sensitivity of @xmath1 , expected to be reached at @xmath2 ghz frequencies in a @xmath3hr integration with the forthcoming alma instrument , virial shocks associated with massive ( @xmath4 ) clusters will stand out from the noise , and can be detected at high significance .
more generally , our results imply that the projected sz surface brightness profile in future , high resolution experiments will provide sensitive constraints on the density profile of cluster gas . |
in principle , relationships between qsos euv continuum and the emission - line gas that it ionizes should give us clues to any relationship between accretion power and the physical conditions and kinematics of accreting or outflowing material within @xmath4 1 pc , hence clues to the mechanism of the central engine .
several studies have shown that the soft x - ray spectrum is related to the emission line spectrum : steeper x - ray spectra are associated with stronger optical feii ( blr ) emission , narrower ( blr ) h@xmath2 , weaker [ oiii]@xmath15007 ( nlr ) emission ( boroson & green 1992 ; grupe 1996 ; grupe et al . 1998 ; forster 1995 ; laor et al .
1994 , 1997 ; corbin 1993 ) .
one of the greatest sources of variation from one spectrum to another can be represented as a linear combination of these observables the so - called `` eigenvector 1 '' of principal component analyses .
its underlying physical cause is unknown , but an understanding seems likely to hold a clue to accretion conditions , and to the energy budget problem in particular the too - great strength of low - ionization emission lines such as feii ( netzer 1985 ) .
laor et al .
( 1997 ) have investigated the soft x - ray and h@xmath2[oiii]@xmath15007 region for a complete sample of all 23 qsos from the pg uv - excess survey ( schmidt & green 1983 ) with @xmath5 , and low galactic absorption ( n@xmath6@xmath7 ) , discovering strong eigenvector 1 relationships in this sample .
the low redshift ensures detection of the soft x - ray emission down to the lowest possible rest frame energy ( typically 0.2 kev ) , which is redshifted into the unobservable soft x - ray region ( @xmath8 ) in higher redshift quasars . the low galactic absorbing column , and accurate 21 cm measurements of this column for all objects ,
ensure small , accurate corrections for ultraviolet and soft x - ray absorption .
this sample is ideal for extending this study into the ultraviolet , where the highest energy continuum and important uv diagnostic lines can be measured with minimal confusion from intergalactic absorption lines .
thus we have obtained hst fos spectrophotometry from wavelengths below ly@xmath0 to beyond the atmospheric cut - off , and mcdonald observatory spectrophotometry from the atmospheric cut - off to beyond h@xmath0 .
instrumental resolutions range from 230 350 km s@xmath9(fwhm ) . here
we present highlights of a first look at our own and archival hst spectra and the x - ray and optical measurements presented by laor et al .
( 1997 ) , and boroson & green ( 1992 ) .
we have measured strengths , ratios and widths ( fwhm ) for the following emission lines : ly@xmath0 with nv@xmath11240 removed , civ@xmath11549 with niv@xmath11486 , heii@xmath11640 and [ oiii]@xmath11663 removed , and we have deblended siiii]@xmath11892 and ciii]@xmath11909 . in most cases it was possible to define a ` rest frame ' wavelength scale referred to [ oiii]@xmath15007 in our mcdonald spectra . generally feiii does not contribute much to the @xmath11909 blend .
evidence for this is that the wavelength of the peak corresponds to within 0.5 - 1 rms of the expected wavelength of ciii]@xmath11909 .
an exception is mkn 478 , where feiii is a clear contributor .
the greatest uncertainties in line measurements arise from uncertainties in continuum placement , and in removal of associated and galactic interstellar absorption .
details will be presented by wills et al .
( 1998b ) .
table 1 presents a few of the correlation results .
eigenvector 1 observables , given in the first column , are correlated with important emission line parameters of the ultraviolet spectrum , given across the top of the table .
eigenvector 1 variables are chosen so as to correlate positively with x - ray spectral index @xmath10 ( f@xmath11 ) .
correlation coefficients are generally pearson coefficients using line ratios , and logarithms of equivalents widths and fwhms .
spearman rank correlations give similar results .
the two - tailed significance levels are given at the end of the table .
we note that a large fraction of our observationally - independent parameters are correlated .
this means that the significance of an individual correlation is not much affected by the fact that we attempted a large number of correlations .
figure1 plots some of the correlations of table1 , the four columns representing eigenvector 1 observables : the steepness of the x - ray spectrum , the strength of feii ( optical ) , the strength of nlr emission ( [ oiii]@xmath15007 ) , and the width of the broad h@xmath2 line . in figure 2
we show the intensity ratio siiii]/ciii ] plotted against feii / h@xmath2 and ly@xmath0/civ .
c@c@c@c@c@c@c & + parameters & @xmath12 & @xmath13 & @xmath14}{\rm c\,iv}$ ] & ew(civ ) & @xmath15}{\rm c\,iii]}$ ] & fwhm ( ciii ] ) + + + @xmath16 & 0.78 & 0.42 & @xmath17 & @xmath180.67 & @xmath17 & @xmath180.59 + ew ( feii ) & 0.69 & 0.70 & 0.53 & @xmath180.68 & 0.72 & @xmath180.52 + @xmath19 & 0.80 & 0.85 & 0.54 & @xmath180.46 & 0.89 & @xmath180.63 + @xmath20}}$ ] & 0.71 & 0.59 & 0.63 & @xmath180.66 & 0.28 & @xmath180.37 + @xmath21}}$ ] & 0.58 & 0.40 & 0.56 & @xmath180.59 & 0.56 & @xmath180.47 + @xmath22}}$ ] & 0.63 & 0.40 & 0.60 & @xmath180.64 & 0.60 & @xmath17 + @xmath23 & 0.68 & 0.65 & 0.53 & @xmath180.69 & 0.56 & @xmath180.78 +
table 1 shows that line ratios involving civ strength , including the ew ( civ ) , correlate significantly with nearly all eigenvector 1 observables in the sense that civ strength anticorrelates with steep soft x - ray spectrum , and strong feii , and correlates positively with [ oiii ] strength and fwhm ( h@xmath2 ) .
we note that correlations of ly@xmath0/civ and ew ( civ ) with @xmath10 are in the same sense as found by wang et al .
( 1998 ) for a large , heterogeneous sample .
our result suggests that eigenvector 1 is correlated with strengths of lines from low ionization transitions ( feii ) and anti - correlated with higher ionization transitions ( [ oiii ] , civ , ovi@xmath11034 ) .
if investigation of species over a wider range of ionization confirms this , one possibile explanation may be along the lines hinted at by boroson & green ( 1992 ) , and investigated more quantitatively by brotherton ( 1996 , 1998 ) : the [ oiii ] may be produced by ionizing photons that reach the nlr after penetrating the blr .
higher blr optical depths will result in stronger emission from lower ionization lines ( mgii@xmath12798 , feii ( optical ) ) , hence reducing the ionizing flux to the nlr beyond .
a more highly ionized blr may produce stronger ovi@xmath11034 , civ@xmath11549 , and lower column densities of mg@xmath24 and fe@xmath24 , allowing greater ionizing flux to reach the nlr .
for example , in a blr consisting of clouds with a distribution of optical depths , increasing ionizing flux could increase the ratio of optically thin to optically thick clouds .
the ratio siiii]@xmath11892/ciii]@xmath11909 is strongly correlated with eigenvector 1 in the sense of a positive correlation with feii ( optical ) , and anticorrelations with [ oiii ] strength and fwhm ( h@xmath2 ) .
it is also correlated with ly@xmath0/civ ( fig .
2 ) . these correlations are stronger than the positive ( negative ) correlations of siiii ] strength ( ciii ] ) alone with feii / h@xmath2 and ly@xmath0/civ .
the similar ( but not identical ) ionization potentials of si@xmath25 and c@xmath25 and photoionization models suggest that these ions are largely cospatial in the blr .
this line ratio is sensitive to density , with n@xmath26 @xmath27 for the siiii]@xmath11892 upper level and n@xmath28 @xmath27 for ciii]@xmath11909 .
thus eigenvector 1 appears strongly related to density , in the sense that the blr high density gas contributes most when feii ( optical ) is strongest .
as found in other samples , fwhm ( ciii ] ) correlates well with fwhm h@xmath2 , suggesting that these are from the same kinematic region .
similarly , there is a good correlation between fwhm ( ly@xmath0 ) and fwhm ( civ ) , but spanning a smaller range of fwhms .
these correlations , together with weaker correlations of fwhm ly@xmath0 with fwhm ( ciii ] ) or fwhm h@xmath2 , and weaker correlations of fwhm ( civ ) with fwhm ( ciii ] ) or fwhm h@xmath2suggest different kinematic origins for low and high ionization gas .
overall , we were somewhat amazed at the strong correlations present in our small sample , and conclude that small , but carefully - defined , homogeneous samples with high quality spectra can yield significant information on the underlying physics .
we suggest that eigenvector 1 , in the sense of increasingly steep soft x - ray spectrum , increasing feii ( optical ) strength , decreasing nlr emission , and narrower ( blr ) h@xmath2 , represents increasing dominance of low - ionization gas , and high densities , as well as decreasing dominance of high - ionization emission .
the interpretation of narrower h@xmath2 lines in terms of higher eddington ratios , and the correlation with steeper soft x - ray spectra , links the accretion process to density and ionization state of the surrounding gas ( wandel , laor , these proceedings ) .
we gratefully acknowledge the help of the following people : c. d. keyes & a. roman of stsci , m. dahlem ( now of estec ) , z. shang , d. r. doss , j. martin , m. villareal , m. ward , d. otoupal , e. green , d. crook , & m. cornell of mcdonald observatory & the university of texas astronomy department . this research is supported by nasa through ltsa grant number nag5 - 3431 ( b.j.w . ) and grant number go-06781 from the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas5 - 26555 .
boller , t. , brandt , w. n. , & fink , h. 1996 , a&a , 305 , 53 boroson , t. a. , & green , r. f. 1992 , , 80 , 109 brotherton , m. s. 1996 , phd dissertation , university of texas at austin brotherton , m. s. 1998 , in preparation corbin , m. r. 1993 , , 403 , l9 forster , k. 1995 , bull .
aas , 187 , 5009 grupe , d. 1996 , ph.d .
thesis , georg - august universitt , gttingen grupe , d. , beuermann , k. , mannheim , k. , thomas , h .- c .
1998 , a&a , submitted laor , a. , fiore , f. , elvis , m. , wilkes , b. j. , & mcdowell , j. c. 1994 , , 435 , 611 laor , a. , fiore , f. , elvis , m. , wilkes , b. j. , & mcdowell , j. c. 1997 , , 477 , 93 netzer , h. 1985 , , 289 , 451 schmidt , m. s. , & green , r. f. 1983 , , 269 , 352 wang , t. , et al .
1998 , a&a , 000 , 000 wills , b. j. , laor , a. , brotherton , m. s. , wills , d. , wilkes , b. j. , & ferland , g. j. 1998a in ` quasars as standard candles for cosmology ' , asp conference series , ed .
ferland , g. j. ( 1998 : asp , san francisco ) .
wills , b. j. , laor , a. , brotherton , m. s. , wills , d. , wilkes , b. j. , & ferland , g. j. 1998b , in preparation .
wills after the meeting i have checked this . while your reported anticorrelation between ew ( civ ) and feii ( optical ) strength may be the same as what we find , i m not sure .
your ew ( civ ) apparently was measured in a way rather different from the conventional one ( marziani et al .
1996 , , 104 , 37 ) and the sample you used was very heterogenous , covering a wide range in luminosity .
our ew ( civ ) measurements include the total line flux , but with niv@xmath11486 , heii@xmath11640 , and oiii]@xmath11663 removed . | a unique , essentially complete sample of 22 qsos , with high quality soft x - ray spectra from rosat , as well as hst and optical spectrophotometry from below ly@xmath0 to above h@xmath0 , is being used to investigate the relationships among the ionizing continuum and the optical and uv continuum , emission and absorption lines . here
we present a first analysis showing that optical ` eigenvector 1 ' linking steeper soft x - ray spectra with increasing optical feii strength , decreasing [ oiii]@xmath15007 emission , and narrower blr h@xmath2 emission , extends to the uv emission lines , and is manifested by weaker civ@xmath11549 emission , stronger siiii]@xmath11892/ciii]@xmath11909 ratio , and narrower ciii]@xmath11909 emission .
steeper soft x - ray spectra have been linked to higher l / l@xmath3 ratios , thus apparently linking blr densities , high and low ionization gas , and kinematics , to the accretion process . |
the efficient market hypothesis ( emh ) has significant influence on theory as well as practical business in the financial literature @xcite .
various evidence concerning market efficiency has been recently discussed .
also the predictability of future price change in the stock market is a very interesting area in the financial field .
research topics on both the degree of efficiency and predictability generally have been known to be intimately related . in a weak - form emh ,
a lower degree of efficiency means that the usefulness of past price change is high in terms of the prediction of future price change .
it is difficult to predict the future price change relatively with a higher degree of efficiency .
however , the relationship between the degree of efficiency and prediction power has not been empirically studied much relatively . in this study
, we investigate the relationship using the stock market indices .
the present study needs a method to quantify the degree of efficiency and prediction power .
the concept of the efficiency corresponds to the weak - form emh concerning whether the information of past price change pattern is useful in terms of the prediction of future price change .
therefore , we employed the quantified measurement for the degree of efficiency and the prediction method using its property directly based on the degree of similarity of price change pattern in the time series . also , we use the hurst exponent to observe the long - term memory @xcite , and approximate entropy ( apen ) to observe the randomness in the time series @xcite . for the quantitative measurement of prediction power , the hit - rate estimated by the nearest - neighbor prediction method ( nn , so called time - delay embedding techniques or a local linear predictor ) is used @xcite . the nn method using reconstructed data which is created by the embedding
is useful to predict the future price change based on the degree of similarity of the price change pattern .
the research topic concerning the existence of long - term memory property in the financial time series is generally well - known in the financial field as well as in other scientific fields .
when a market has a long - term memory property , the market does not reflect the existing information both immediately and sufficiently .
therefore , if financial time series has a long - term memory property , the information of the past price change becomes a valuable one to predict the future price change .
the long - term memory property of the return and volatility in the stock markets has been an active research field @xcite .
the hurst exponent , a measurement of the long - term memory property , quantify the degree of market efficiency of the stock market @xcite .
however , the empirical evidence on whether the results of degree of efficiency for long - term memory property directly relates to the prediction power of future price changes based on the patterns of past price changes has not been suggested yet . in this study
, we empirically investigate the relationship between the values and prediction power of future price change direction based on the hurst exponent to quantify the degree of efficiency of stock market .
the apen is also a measurement to quantify degree of efficiency in the financial time series , which can quantitatively estimate the degree of randomness in the time series .
it calculates quantitatively complexity , randomness , and prediction power .
let us consider the similarity of price change pattern in the financial time series .
when the frequency of the similarity pattern is high , the randomness of the time series is low and the apen also low .
however , the apen has a high value if the frequency is low .
previous works which introduce and apply the apen measurement to the financial data @xcite and argue that the apen value has significant information in terms of measuring degree of efficiency using the foreign exchange data @xcite are announced .
according to the previous studies , the correlation between the hurst exponent and apen value is negative @xcite .
however , the evidence on whether the results of degree of efficiency due to randomness directly relates to the prediction power of future price changes based on the patterns of past price changes has not been suggested yet . in this study
, we investigate empirically the relationship between the values and prediction power of future price change based on the apen to quantify the degree of efficiency of stock markets .
we also study a prediction method which directly uses the similarity pattern of past price change as the prediction of future price .
we employ the nn method to predict future price change based on the similarity of price change pattern . according to the results of previous studies ,
the nn method is useful to predict the financial time series within a short time frame @xcite . in this study
, we also investigate the relationship between the prediction power estimated by th nn method and the hurst exponent , the apen , respectively .
this investigation observes the relationship between prediction power and degree of efficiency in a financial time series .
the prediction power that we use in this paper is a hit - rate , which quantify the consistency between the actual price change and the predicted one by the nn method . according to the established research purpose and previous studies
, we originally expected a positive relationship between the hurst exponent and prediction power .
in other words , a market index with a higher hurst exponent shows averagely a higher prediction power than one with a lower hurst exponent .
also , we expected a negative relationship between the apen value and prediction power ; a market index with a lower apen value has averagely a higher prediction power than one with a higher apen . through this study , we find the following results .
first , the relationship between the average hurst exponent @xmath7 as the long - term memory property and the average apen value @xmath8 as the randomness is negative , @xmath9 where @xmath1 represents cross - correlation and @xmath10 each stock market .
second , the average prediction power @xmath11 in terms of the direction of future price changes reveals a positive relationship with the hurst exponent @xmath12 , and has a negative one with the apen @xmath13 . according to the above results , we find that the relationship between the degree of efficiency and the prediction power is negative . in other words ,
a market index with a lower degree of efficiency has a higher predictability of the future price change in terms of the past price change pattern than one with a higher degree of efficiency .
moreover , we find that the hurst exponent as the measurement of long - term memory property provides valuable information in terms of the prediction of future price change . in the next section , we describe the data and methods of the verification process used in this paper , in section [ sec : results ] , we present the results obtained according to the established research purpose .
finally , we summarize the findings and conclusions of the study .
we study the daily indices of 27 stock markets , which obtained from datastream ( _ http://www.datastream.net_ ) .
the market indices are composed of 13 markets in asia - pacific region , 7 in north and south america , and 7 in europe .
they are china ( shenzhen composite ) , hong kong ( hangseng ) , india ( bombay se200 ) , indonesia ( jakarta se composite ) , japan ( nikkei225 ) , korea ( kospi200 ) , malaysia ( kuala lumpur composite ) , philippines ( se composite ) , singapore ( straits times ) , taiwan ( se weighted ) , thailand ( bangkok s.e.t ) , australia ( asx ) , new zealand ( nzx ) , argentina ( base general ) , brazil ( bovespa ) , chile ( general ) , mexico ( ipc ) , peru ( lima se general ) , canada ( tsx composite index ) , usa ( s&p 500 ) , austria ( atx ) , denmark ( copenhagen kbx ) , france ( cac 40 ) , germany ( dax 30 ) , italy ( milan comit general ) , netherlands ( aex index ) , and the uk ( ftse100 ) .
the data period is 15 years from january 1992 to december 2006 except australia ( from june 1992 ) , argentina ( from june 2000 ) , and denmark ( from january 1996 ) .
the return time series of the market index used in this paper was calculated by the logarithmic change of the price , @xmath14 where @xmath15 is the stock price on day @xmath16 . in the previous studies , researchers utilized various methods such as the hurst exponent @xcite , arfima ( autoregressive fractional integration moving average ) @xcite and figarch ( fractionally integrated generalized autoregressive conditional heteroscedasticity ) @xcite to quantitatively observe the long - term memory property in a financial time series .
however , the long - term memory property in a financial time series does not depend on the method used .
we use the hurst exponent to quantify the degree of long - term memory property .
there are many methods to calculate the hurst exponent .
they are the classical re - scaled range analysis @xcite , generalized hurst exponent method @xcite , modified r / s analysis @xcite , gph method @xcite , and the detrended fluctuation analysis ( dfa ) @xcite .
according to the previous works , the dfa method is the most efficient method for accurately calculating the hurst exponent @xcite .
therefore , we select the hurst exponent calculated by the dfa method as a measurement to quantify the degree of efficiency in a financial time series .
the hurst exponent calculated by the dfa method can be explained by the following steps .
first , after the subtraction of the mean @xmath17 from the original time series @xmath18 , one accumulates the series @xmath19 as defined by @xmath20 , \tag{1a}\end{gathered}\ ] ] where @xmath21 is the number of the time series .
next , the accumulated time series is divided into a window of the same length @xmath22 .
we estimate the trend lines @xmath23 by using the ordinary least square ( ols ) in each window .
we eliminate trends existing within the window by subtracting from the accumulated time series in each window .
this process is applied to every window and the fluctuation magnitude @xmath24 is defined as @xmath25 the process mentioned above is repeated for every scale @xmath26 . in addition , we investigate whether the scaling relationship exists for every scale .
that is , the scaling relationship is defined by @xmath27 where @xmath28 is the constant and @xmath2 is the hurst exponent . if @xmath29 , the time series has a short - term memory . if @xmath30 , it has a long - term memory .
we use the hurst exponent as the measurement to quantify the degree of market efficiency of a stock market in each market . as the hurst exponent increases ,
the persistence of similarity patterns in the past price changes is high . in other words ,
the pattern of past price change is a valuable information when predicting the patterns of future price change .
in addition , in order to estimate and use the robust hurst exponent regardless of the time variation , we use the average hurst exponent,@xmath31 , estimated repeatedly until december 2005 @xmath32 by estimating the time window with the width of 5 years and shifting 1 year for a whole period until december 2006 @xmath33 .
we omit the data of the year 2006 when calculating the mean hurst exponent because an out - of - sample of the 1-year prediction period is required in the nn method .
the apen , which measures the degree of randomness in the time series , was also introduced as the measurement of the degree of efficiency of the financial time series @xcite .
we use the apen as the second measurement of the degree of efficiency in this study .
the apen is defined as @xmath34 where @xmath35 is the embedding dimension and @xmath36 is the tolerance to determine the similarity between price change patterns .
the @xmath37 is given by @xmath38 , \tag{2b } \\
c_{i}^{m } ( r ) = \frac{\displaystyle b_{i}(r)}{(n - m+1 ) } , \tag{2c}\end{gathered}\ ] ] where @xmath39 is the number of data pairs within a tolerance of similarity @xmath36 . also , we calculate the similarity in the time series of each price change pattern @xmath40 by the distance @xmath41 $ ] between two vectors @xmath42 defined as @xmath43 \leq r. \tag{2d } \\ d[x(i),x(j)]= \underset{k=1,2,
.. ,m}{max}(|u(i+k-1)-u(j+k-1)| ) . \tag{2e}\end{gathered}\ ] ] through the above equations , the apen compares the relative magnitude between repeated pattern occurrences for the embedding dimensions @xmath35 and @xmath44 .
in other words , the apen becomes smaller as the frequency of similar price change patterns for the embedding dimension @xmath35 equals that for the embedding dimension @xmath44 . as both frequencies are same , the apen value is 0 . on the other hand , apen is relatively larger as long as the frequency of similar price change patterns for the embedding dimension @xmath44 is small . as the apen is small ,
the frequency of the similar price change pattern is small , the time series data has a low degree of randomness , and the efficiency of financial time series becomes low . in order
to calculate the apen , we use the embedding dimension , @xmath45 , and tolerance of similarity , @xmath46 of standard deviation of the time series , similar to previous works @xcite .
we establish the estimated period to calculate the apen as the same process for the hurst exponent . in order to use the apen robustly
regardless of time variation , we use the average apen value , @xmath47 , estimated repeatedly until december 2005 @xmath48 by estimating a period of 5 years and shifting 1 year for the whole period until december 2006 @xmath33 .
the reason we omit the data during 2006 when calculate the mean apen is that we need an out - of - sample of the 1-year prediction period not included in the estimated period of the nn prediction method .
in this section , we introduce the nn prediction method , which investigates the relationship between degree of efficiency as the hurst exponent or the apen and the predictability of future price change . in order to observe the relationship between the degree of efficiency and predictability , a prediction method that uses a similar pattern of past price changes
should be used as we used the information of similar patterns in the price change to calculate the hurst exponent and apen value . for this reason
, we used the nn prediction method .
many previous studies have confirmed the usefulness of the nn prediction method @xcite , and the process of the nn prediction method can be expressed as follows . in the first step
, we reconstruct the pattern series , @xmath49 , with the embedding dimension , @xmath35 , and the time delay , @xmath50 from the financial time series , @xmath51 and defined as @xmath52 \tag{3a } \\
n = ( m-1)\tau+1,\ldots , t \notag\end{gathered}\ ] ] the structure of the reconstructed time series can be explained as follows .
the length of the reconstructed pattern series @xmath53 is @xmath35 in a time series of @xmath21 length , and this creates the @xmath54 data points . in order to predict price change at time @xmath55
, we have to find price change patterns similar to the target price change pattern @xmath56 at time @xmath16 among a @xmath54 past price pattern series @xmath53 . the criteria to select the similar patterns in the price change
is the distance between both vectors .
@xmath57 when the distance is 0 , both price change patterns are completely identical . as the distance increases , they do not have similar patterns of price change .
we choose the @xmath58 number of similar pattern @xmath59 , which has the smallest distance , among the whole pattern series . to confirm whether the chosen similar patterns of past price changes are true , we select the chosen similar patterns @xmath60 for the embedding dimension , @xmath44 . the future price change @xmath61 can be predicted by the next price change @xmath62 of the chosen patterns of the price change @xmath63 and defined as @xmath64 we calculate the hit - rate by the degree of consistence as the prediction power by comparing the direction of actual price change with direction of predicted price change . in addition , in order to employ the instinctive method in the estimation of prediction power , we determine the direction by comparing it to the frequency of direction of past price change patterns .
that is , the direction of future price @xmath61 is determined by the direction having a higher frequency using the number of direction frequency on whether the next day s price change of the chosen past patterns in the eq .
( 3c ) is in an upward direction or downward direction .
we use the same period described above to calculate the hurst exponent and the apen . in other words , the period to reconstruct is 5 years .
the prediction period is 1-year of out - of - sample , not including the reconstruction period .
the prediction day is one day . that is , using the first reconstruction period @xmath65
, we predict the direction ( up or down ) of price change for the first trading day @xmath66 in the prediction period ( one year ) . and
, we valuate the consistency between actual price change and that estimated by the nn model . in order to predict the direction of price change on the second trading day @xmath67
, we use the pattern series by removing the farthest trading day ( 1 day ) and adding the new trading day ( 1 day ) .
this process iss repeated for the prediction period ( 1 year ) @xmath68 . after the prediction period closes , the prediction power can be calculated by the frequency ratio @xmath69^{(\ast)}]/fq_{t}[t_p]])$ ] of the number of consistent direction trading @xmath70 with the number of trading day @xmath71 for the prediction period .
next , according to the above process after the moving period ( 1 years ) as the process for the hurst exponent and the apen , we calculate the hit - rate @xmath72 of future price change using the second reconstruction period @xmath73 .
this process is repeated until the reconstruction period @xmath32 of the last day of the prediction period .
we use the average hit - rates @xmath74 in terms of the consistent direction between actual price changes and those estimated by the nn model as the prediction power to observe the relationship with the hurst exponent and apen as the degree of efficiency .
we present the observed relationship between the hurst exponent and the apen to measure the degree of efficiency and the relationship between degree of efficiency and prediction power . in this section , using the stock market indices of 27 markets , we present the investigated results for the relationship between the hurst exponent and the apen value in terms of degree of efficiency . as mentioned above , we expected a negative relationship between the hurst exponent and apen . in other words , from the viewpoint of the hurst exponent , a lower efficiency degree means a higher hurst exponent which reveals long - term memory .
on the other hand , a lower efficiency degree concerning the apen means lower apen values , which indicate a lower degree of randomness .
of course , based on both measurements , a lower efficiency degree occurs when similar patterns occurs repeatedly in the past price change .
futhermore , the observed results are negative evidence of the weak - form emh . the results are presented in fig .
1 . fig .
1 shows the average hurst exponent @xmath75 and the apen @xmath76 calculated repeatedly by using the estimated period ( 5 years ) and moving period ( 1 year ) for each stock indices @xmath77 . in fig .
1 , the x - axis and y - axis indicate the apen and hurst exponent , respectively . fig .
1(a ) and ( b ) show the results for using a time series of the actual stock market and using a random time series with the mean and standard deviation of time series of the actual stock indices , respectively .
in the case of a random time series , the hurst exponent and apen also have an average value @xmath78 for the whole period .
the market index of each country denotes according to a continent .
that is , the circles ( red ) , triangles ( magenta ) , squares ( blue ) , and diamonds ( green ) indicate the asia - pacific region ( 13 countries ) , south and north america ( 7 countries ) , and europe , respectively . according to the results , we find that the relationship between the hurst exponent and the apen is negative @xmath79 $ ] . however , we could not find these relationships in the random time series .
that is , the apen in terms of the randomness of the price change patterns is low as the financial time series has a high long - term memory property .
therefore , both methods can be used complementarily in terms of measuring the degree of efficiency in a financial time series .
as the market indices locate at the top left , they have either a higher long - term memory or lower randomness .
this shows a low degree of efficiency in the time series .
we discovered that most stock indices for south america or asia belong to this area . on the other hand ,
as the market indices locate at the bottom - right , they have lower long - term memory or higher randomness .
that is , this demonstrates a high degree of efficiency .
most stock indices in this position are in europe ( france , the uk , germany ) , north america ( canada , the usa ) , and asia ( australia , japan , new zealand and korea ) .
in addition , we investigated whether the degree of efficiency increases more in the past period in each market indices .
the results are presented in fig .
2 . based on the hurst exponent and apen estimates using each sub - period @xmath80 including 2006 , we observed measurements at times @xmath81 and @xmath82 .
the first sub - period @xmath65 for the stock indices is 5 years , from january 1992 to december 1996 , except for three countries , australia ( june , 1992 ) , argentina ( june , 2000 ) , and denmark ( january , 1996 ) .
the last sub - period @xmath33 is 5 years , from january 2002 to december 2006 .
2(a ) and ( b ) show the apen @xmath83 and hurst exponents @xmath84 of the market indices of 27 countries .
2 , the first ( @xmath81 , the past ) and last ( @xmath82 , the present ) sub - periods are denoted by squares ( blue ) and circles ( red ) . by the results , on the whole , we find that the degree of efficiency in market indices increase nowadays more than in past periods . in other words ,
the hurst exponent of asia decreased from 0.56 to 0.51 , and south - america decreased from 0.61 to 0.57 , respectively .
the apen for asia increased from 1.59 to 1.65 , for south - america from 1.56 to 1.63 . on the other hand ,
the hurst exponent and apen of stock indices for europe reveal a different result .
that is , the hurst exponent of past and present approaches 0.5 averagely , but the apen decreases averagely from 1.69 to 1.53 .
we could not find consistent changes in terms of the degree of efficiency . through the results
, we confirmed that both methods can be a measure of the degree of efficiency of a financial time series .
in addition , we can complementarily use both methods as the standard index of the degree of efficiency because the relationship between both methods is negative . although we could not find consistent results between apen measurement and hurst exponents for europe , the present efficiency degree of the market index for each country increased more than those of past periods .
in this section , we present the results of the empirical relationship between degree of efficiency and prediction power of future price changes estimated by the nn prediction method . generally , the predictability of a time series having a lower degree of efficiency is higher than that of a time series having a higher degree of efficiency .
therefore , the expected relationship between the hurst exponent and apen values in terms of the degree of efficiency and prediction power estimated by the nn prediction method is as follows .
we expect a positive relationship between the hurst exponent and prediction power of the nn prediction method .
that is , a market index having a higher hurst exponent value has a high long - term memory property because there are continuously similar patterns of past price change .
therefore , the predictability of future price change using the patterns of the past price changes when the market index increases .
however , we expect a negative relationship between apen and prediction power . that is , a market index having a lower apen has a lower randomness because the similar patterns of past price changes happen frequently . therefore , the predictability of the future price change using the patterns of the past price changes as the market index increases .
the results are presented in fig .
3 . in fig .
3 , the prediction power of the nn prediction method is an average value @xmath85 of the hit - rates in terms of the direction of future price changes in each sub - period . fig .
3(a ) shows the relationship between average apen @xmath86 and average prediction power @xmath87 and figure 3(c ) shows the relationship between average hurst exponent @xmath88 and average prediction power @xmath87 .
moreover , using the random time series created with the mean and standard deviation of the original market index , the relationship between nn prediction power and apen or hurst exponent is presented in figs .
3(b ) and 3(d ) , respectively .
the market index of each country is denoted according to continent and color .
the asia - pacific region , south and north america , and europe are denoted by circles(red ) , triangles(magenta ) , squares(blue ) , and diamonds ( green ) , respectively . according to the results
, we can confirm the relationship between degree of efficiency and prediction power , generally acknowledged in the financial literature .
that is , in fig .
3(a ) , the relationship between average hit - rates estimated by the nn method and average apen is negative @xmath89 , while in fig.3(c ) , the relationship between average hit - rates and average hurst exponent shows a strong positive relation @xmath90 .
therefore , stock indices with a lower degree of efficiency ( higher hurst exponent or lower apen ) have a higher prediction power in terms of the direction of future price change than those with the higher degree of efficiency .
of course , we could not find figs .
3(b ) and 3(d ) using the random time series .
the places which have a lower degree of efficiency and higher predictability are shown in the top left of fig .
3(a ) , which displays a relationship between apen and prediction power and the top right of fig .
3(c ) , which shows a relationship between the hurst exponent and prediction power .
these market indices are generally found in south america and asia .
however , places which have a higher degree of efficiency or lower predictability are shown in the bottom right of fig .
3(a ) and the bottom left of fig .
the market indices displayed here are found in europe ( france , the uk , germany _ et al .
_ ) , north america ( the usa ) , and asia ( australia , japan , new zealand _
et al . _ ) .
according to the above results , we discovered that the degree of efficiency of a financial time series is significantly related to the prediction power of future price changes .
we also found that the hurst exponent and apen provide significant information in terms of the nn prediction method using the similarity of past price change patterns .
in particular , we found that the hurst exponent used to measure long - term memory property is closely correlated with the prediction power of the nn prediction method .
we empirically investigated the relationship between the degree of efficiency and the prediction power using the market index for various countries . according to the viewpoint of the weak - form emh ,
a lower degree of efficiency means that the information of the past price changes is useful in terms of the predictability of future price change , and a higher efficiency degree demonstrates that the predicting is difficult relatively .
the efficiency in this paper is based on the weak - form emh which used to show whether the information of past price changes patterns is useful in terms of the prediction of future price changes .
we employed the measurements such as the hurst exponent and the apen which quantify the degree of efficiency and the prediction method ( nn prediction method ) by using the measurements .
the measurements and the method are based on the similarity of price change pattern in the time series .
we summarize the results as follows .
first , we found that the relationship between the hurst exponent and the apen , the measurements of the degree of efficiency , is negative .
therefore , the hurst exponent and the apen are useful to measure the degree of efficiency in the financial time series complementally .
second , we found that the degree of efficiency is positively related to the prediction power .
market indices having a lower degree of efficiency ( high hurst exponent or low apen ) have a higher prediction power in terms of the direction of the future price change than those having a higher degree of efficiency ( low hurst exponent or high apen ) .
futhermore , we found that the hurst exponent has a very strong correlation with the prediction power of nn prediction power . according to the above results , we _ empirically _ confirm the relationship between the degree of efficiency and the prediction power , which have acknowledged _ qualitatively _ in the financial field .
we also found that the hurst exponent and the apen provide useful information to the nn prediction method using the similarity of past price change pattern .
in addition , our study provides useful information in terms of the consistency of the international funds which have considerable attention in the financial field and the selection of active and passive investment countries as investment strategy , and the investment weights for future works .
also , further works as the complementary process are expected . | this study investigates empirically whether the degree of stock market efficiency is related to the prediction power of future price change using the indices of twenty seven stock markets .
efficiency refers to weak - form efficient market hypothesis ( emh ) in terms of the information of past price changes .
the prediction power corresponds to the hit - rate , which is the rate of the consistency between the direction of actual price change and that of predicted one , calculated by the _
nearest neighbor prediction method _ ( nn method ) using the out - of - sample . in this manuscript ,
the _
hurst exponent _ and the _ approximate entropy _ ( apen ) are used as the quantitative measurements of the degree of efficiency .
the relationship between the hurst exponent , reflecting the various time correlation property , and the apen value , reflecting the randomness in the time series , shows negative correlation ( @xmath0 where @xmath1 represents cross - correlation , @xmath2 the hurst exponent , and @xmath3 apen ) .
however , the average prediction power on the direction of future price change has the strongly positive correlation with the hurst exponent ( @xmath4 where @xmath5 represents the average prediction power calculated by the nn method ) ) , and the negative correlation with the apen ( @xmath6 ) .
therefore , the market index with less market efficiency has higher prediction power for future price change than one with higher market efficiency when we analyze the market using the past price change pattern .
furthermore , we show that the hurst exponent , a measurement of the long - term memory property , provides more significant information in terms of prediction of future price changes than the apen and the nn method . |
the reader should note that this is an identical copy of an article first published in @xcite .
beam effects have limited the performance of previous and existing hadron colliders @xcite such as the isr @xcite , sp@xmath0s @xcite , tevatron @xcite and rhic @xcite , and are also expected to limit the performance of the lhc @xcite .
beam beam effects can be categorized as either incoherent ( dynamic aperture and beam lifetime ) , pacman ( bunch - to - bunch variations ) , or coherent ( beam oscillations and instabilities ) @xcite .
these effects can be caused by both head - on and long - range interactions .
head - on effects , leading to tune shifts and spreads , are important in all hadron colliders .
total beam beam induced tune shifts as large as 0.028 were achieved in the sp@xmath0s @xcite and tevatron @xcite , although operational tune shift values are somewhat lower .
the lhc in its early stages of commissioning has already reached a total head - on beam beam tune shift of 0.02 @xcite .
long - range effects , however , differ in previous and existing colliders . in the isr the beams collided under a large crossing angle of 15 deg @xcite that greatly reduced long - range effects . in the sp@xmath0s , with both beams in the same aperture and only three bunches per beam
, there were a few long - range interactions distributed over the ring circumference . due to the difference in the bunch intensities , the effect on the anti - protons was stronger . in the tevatron , also with both beams in the same aperture but 36 bunches per beam , there are more long - range interactions . with increased intensity of the anti - proton bunches , protons can also be affected . in rhic ( fig .
[ fig:01 ] ) , where both beams share the same aperture only in the interaction regions , there is only one long - range interaction per interaction region without an experiment ( a total of four in the current configuration ) , with a 10 mm separation ( corresponding to 30 rms beam sizes for protons at 250 gev energy ) .
long - range interactions have affected the rhic ramp transmission in the past @xcite .
in the lhc there are 32 long - range beam beam interactions localized in each of four interaction regions @xcite .
a major upgrade of the lhc interactions region is foreseen by the end of the decade with the primary objective to increase the average luminosity of the machine by about a factor of 5 to 10 above the design performance . among the various upgrade scenarios a crab crossing scheme ( cc ) , an early beam separation scheme ( es ) , and a large piwinski angle ( lpa )
are considered . in the cc scheme , crab cavities placed on either side of the interaction region
impart a transverse kick to effectively compensate the crossing angle .
this scheme allows for a large crossing angle that greatly reduces long - range beam beam effects . in the es scheme @xcite the number of long - range interactions is greatly reduced but four parasitic collisions at 45 @xmath1 per ip remain . in the lpa scheme @xcite the small crossing angle will be maintained , and long bunches of intensities up to @xmath2@xmath3 protons are used .
all schemes aim at higher than nominal bunch currents and reduced @xmath4 .
therefore , long - range effects tend to become more problematic and require more aperture for larger crossing angles or compensation to mitigate these effects .
the lpa scheme would most benefit from long - range beam beam compensation .
the compensation of long - range effects in the tevatron was proposed with electron lenses @xcite , and in the lhc with wires @xcite .
electron lenses were also considered for the lhc @xcite , and the use of wires was also studied for the tevatron @xcite .
implementation of long - range beam beam compensation in the tevatron is challenging because the effect is distributed over the whole ring . in the lhc
the effect is localized in the interaction regions .
a partial long - range beam beam compensation was successfully implemented in the @xmath5 collider da@xmath6ne @xcite .
beam compensation and related issues were reviewed at a workshop in 2007 @xcite .
figures [ fig:01 ] and [ fig:02 ] show the basic layout of the beam beam interaction and compensation studies in rhic . at store
there are nominally two head - on interactions in points 6 and 8 ( ip6 and ip8 ) , and long - range interactions with a large separation in the other interaction points .
three bunches in the blue ring are coupled to three bunches in the yellow ring through the head - on beam beam interaction . for studies ,
two dc wires were installed in the blue and yellow rings respectively in interaction region 6 ( ir6 ) .
table [ tab : rhicbb ] shows the main beam parameters for polarized proton operation , both achieved and design . in rhic the beam beam effect is strongest in proton operation .
.main rhic parameters achieved in polarized proton operation that are relevant for beam beam effects protons ( 2009 ) .
note that the polarized proton bunch intensity is also limited by intensity dependent depolarization effects in the ags . [ cols="<,^,^,^",options="header " , ] prior to a long - range compensation attempt , a position scan of the wire on each beam was performed with a wire current of 50 a. a 36@xmath736 bunch pattern with six non - colliding bunches was chosen to enable a comparison of the lifetime in the presence of the wire between single beam and colliding beams simultaneously .
the corresponding beam loss rates as a function of beam to wire separation on both colliding and non - colliding bunches were measured .
the initial beam loss rates with colliding beams were stabilized to the nominal 10% per hour .
the maximum total beam losses for the wire movements towards the beam at fixed current were constrained to 100150% per hour for a very short period to avoid disrupting the beam quality significantly for subsequent measurements .
figure [ honoho ] shows the evolution of the intensity between bunches with and without head - on collisions .
it is evident that the bunches with the head - on collisions have a more severe effect from the lr forces of the wire .
several hypotheses can be formulated to explain the increased losses for bunches with head - on collision . *
the dynamic aperture for the bunches with head - on is significantly smaller than that of the single beam which could lead to the observed beam losses .
* it was also suggested by an anonymous referee that the addition of the head - on collisions enhances the diffusion leading to enhanced losses in the presence of long - range interactions .
figure [ honoho ] clearly shows a larger initial slope for bunch intensities with collisions .
however , it is difficult to untangle the contribution from the reduced dynamic aperture as opposed to enhanced diffusion .
* the additional tune shift due to the wire along with large head - on tune shift could lead to beam losses due to very limited tune space available .
no tune optimization was performed during the experiment . *
the effect of the wire on the orbit can introduce a static offset between the two beams at the ip which is approximately proportional to the wire distance .
a large offset due to the kick from the wire can lead to emittance blow - up and beam losses @xcite .
the relative offset at the collision point during the wire scan with 50 a ( see fig . [ orbitoff ] ) is well below the 1 @xmath1 level which is very small .
however , simulations to support each of the above hypotheses to explain its contribution towards observed losses is beyond the scope of this paper . the bunch spacing and the interaction region geometry in rhic does not inherently have lr beam beam interactions .
it is therefore necessary to shift the collision point towards the dx magnet closest to the dc wires as noted before .
this location enables an artificially induced lr interaction between the two beams and simultaneously allows for a minimum phase advance between the lr interaction and dc wires ( 6 deg ) .
additionally , this location has sufficient aperture for an orbit scan with the range of interest ( 310 @xmath1 )
. figure [ fig : orbit ] shows the trajectories of the blue and yellow rings with the lr interaction set at approximately 3.1 @xmath1 .
the individual bunch intensities and beam losses were recorded during the position scan with the lr compensation @xcite .
figure [ lonolo ] shows the beam losses as a function of the wire position . in the blue ring
, the losses are always increasing as the wire approaches the beam .
therefore , no evidence of compensation of the lr interaction from the blue beam is visible .
however , in the yellow ring , the beam lifetime improved as the beam to wire distance approaches 3 @xmath1 ( fig . [ lonolo ] ) .
consecutive retractions and restoration of the beam to wire distance to 3 @xmath1 show similar improvement of the beam lifetime . this indicates a compensation of the effect of lr interaction by the dc wire .
in addition to beam losses , the individual bunch intensities with and without lr interactions and simultaneous compensation is shown in fig .
note that all 36 bunches experience the effect of the dc wire , but only 30 bunches experience lr interactions .
therefore , only bunches with an lr interaction can experience a compensation . in the blue ring ,
the bunch intensity evolution is similar for bunches with and without lr compensation .
hence , only the effect from the wire is visible .
the bunches with lr interaction and simultaneous compensation have reduced beam losses as compared to the bunches that only see the wire .
this is consistent with the beam loss measurements ( fig .
[ lonolo ] ) .
long - range beam beam experiments were conducted in rhic from 2005 to 2009 .
the motivation for these were two - fold .
first , the experimental data can benchmark simulation codes for situations of strong localized long - range beam beam interactions as they will exist in the lhc .
second , the compensation of a single long - range beam
beam interaction can be tested in a scheme that is also usable in the lhc .
these experiments complement the experience with long - range beam beam interactions in the sp@xmath0s and tevatron , wire experiments in the sps , and the partial long - range compensation in da@xmath6ne .
the rhic wires created strong localized long - range beam beam effects , comparable in strength to the effect expected in the lhc , with a beam that has a lifetime typical of hadron colliders , and including head - on beam beam collisions .
the observed orbit and tune changes due to the wire were as expected .
the effect of the long - range beam beam interactions on the beam loss rate is sensitive to a number of beam parameters such as the tunes and chromaticities .
fitting the beam lifetime @xmath8 to an exponential function @xmath9 as a function of the distance @xmath10 between the beam and the wire , exponents @xmath11 in the range between 1.7 and 16 were found .
distances smaller than 5 @xmath1 created losses too large for collider operation .
the experimentally observed distance from the wire to the beam at which large beam losses set in could be reproduced in simulations within 1 @xmath1 .
the beam lifetime with long - range interactions created by the wire was degraded further through head - on collisions . a single attempt to compensate long - range beam
beam interaction via a dc wire showed evidence of compensation .
for discussions and help we are most thankful to collaborators within the us lhc accelerator research program as well as several people at various laboratories . among these are o. brning , r. demaria , u. dorda , w. herr , a. kabel , h .- j .
koutchouk , c. milardi , k. ohmi , t. pieloni , j. qiang , f. schmidt , t. sen , g. sterbini , and f. zimmermann .
99 n. abreu , r. calaga , w. fischer and g. robert - demolaize , phys . rev .
beams 14 ( 2011 ) 091001 .
e. keil , `` beam beam effects in electron and proton colliders , '' pac89 , illinois , march 1989 , pp .
17311735 ( 1989 ) . y. alexahin , `` theory and reality of beam beam effects at hadron colliders , ''
pac05 , tennessee , may 2005 , pp . 544548 ( 2005 ) . j .-
koutchouk , `` a summary on care - hhh_apd presentations on beam beam effects & beam beam compensation at the lhc , '' final care - hhh workshop on scenarios for the lhc upgrade and fair - hhh-2008 - chavannes - de - bogis , switzerland ; w. scandale and f. zimmermann ( eds . ) , cern-2009 - 004 ( 2009 ) ; care - conf-08 - 032-hhh ( 2008 ) . m. jacob and k. johnson ( editors ) , `` the role of the isr in the development of particle physics , '' cern yellow report 84 - 23 ( 1984 ) .
hmery , a. hofmann , j .-
koutchouk , s. myers , and l. vos , `` investigation of the coherent beam beam effects in the isr , '' ieee trans . of nucl .
ns-28 , no . 3 ( 1981 ) .
evans and j. gareyte , `` beam beam and single beam effects in the sps proton - antiproton collider , '' pac83 , santa fee , march 1983 , pp . 23972399
l. evans , j. gareyte , m. meddahi , and r. schmidt , `` beam beam effects in the strong - strong regime at the cern - sps , '' pac89 , illinois , march 1989 , pp .
14031405 ( 1989 ) . k. cornelis , m. meddahi , and r. schmidt , `` the beam beam effect in the sps proton antiproton collider with unequal emittances , '' epac90 , nice , june 1990 , pp .
16701672 ( 1990 ) .
k. cornelis , `` beam beam effects in the sps antiproton - proton collider , '' lhc99 , geneva , june 1999 , cern - sl-99 - 039 ap ( 1999 ) .
x. zhang , t. sen , v. shiltsev , m. xiao , y. alexahin , f. schmidt , and f. zimmermann , `` experimental studies of beam beam effects in the tevatron '' , pac03 , oregon , 2003 , pp .
17571759 ( 2003 ) .
v. shiltsev , y. alexahin , v. lebedev , p. lebrun , r. s. moore , t. sen , a. tollestrup , a. valishev , and x. l. zhang , phys . rev .
beams 8 , ( 2005 ) 101001 .
a. valishev , `` simulation of beam beam effects and tevatron experience '' , epac08 , genoa , 2008 , pp .
29372941 ( 2008 ) .
w. fischer et al . , `` observation of strong - strong and other beam beam effects in rhic , '' pac03 , oregon , 2003 , pp .
135137 ( 2003 ) . v. ptitsyn et al . , `` rhic performance with polarized protons in run-6 , '' epac06 , edinburgh , 2006 , pp .
592594 ( 2006 )
. w. herr , `` computer simulation of the coherent beam beam effect in the lhc , '' pac91 , san francisco , may 1991 , pp .
10681070 ( 1991 ) .
n. gelfand , c. johnstone , t. sen , and w. wan , `` effect of the beam beam interactions on the dynamic aperture of the lhc at collision , '' pac99 , new york , march 1999 , pp .
16771679 ( 1999 ) . j. gareyte , `` beam beam design criteria for lhc , '' lhc99 , geneva , june 1999 , cern - sl-99 - 039 ap ( 1999 ) .
furman and w.c .
turner , `` beam beam simulations for separated beams in the lhc , '' epac00 , vienna , june 2000 , pp .
11961189 ( 2000 )
. h. grote , `` self - consistent orbits with beam beam effect in the lhc , '' epac00 , vienna , june 2000 , pp .
12021204 ( 2000 ) .
leunissen , h. grote , and f. schmidt , `` lhc dynamic aperture including the beam beam force , '' epac00 , vienna , june 2000 , pp .
12081210 ( 2000 ) .
zorzano and t. sen , `` emittance growth of the lhc beam due to the effect of head - on beam beam interaction and ground motion , '' epac00 , vienna , june 2000 , pp .
12261228 ( 2000 ) .
zorzano and f. zimmermann , `` simulations of coherent beam beam modes at the lhc , '' epac00 , vienna , june 2000 , pp .
12291231 ( 2000 )
. w. herr , `` tune spreads and shifts due to the long - range beam beam effect in the lhc , '' epac00 , vienna , june 2000 , pp .
16301632 ( 2000 ) . o. brning , p. collier , p. lebrun , s. myers , r. ostojic , j. poole , and p. proudlock ( editors ) , cern-2004 - 003 ( 2004 ) .
f. zimmermann , `` lhc upgrade scenarios , '' pac07 , albuquerque , june 2007 , pp .
714718 ( 2007 ) .
koutchouk , `` a concept for the luminosity upgrade ... , '' pac07 , albuquerque , june 2007 , pp .
33873389 ( 2007 ) . g. sterbini and g. tommasini , `` a feasibility study of superconducting dipole for the early separation scheme of slhc , '' epac08 , genoa , june 2008 , pp .
24612463 ( 2008 ) .
k. ohmi , `` study of beam beam effect at various collision schemes in lhc , '' epac08 , genoa , june 2008 , pp . 25932595 ( 2008 ) .
u. dorda , `` compensation of long - range beam beam interaction at the cern lhc , '' phd thesis , vienna university of technology , austria ( 2008 ) .
t. pieloni , `` a study of beam beam effects in hadron colliders with a large number of bunches , '' phd thesis , ecole polytechnique federale de lausanne , switzerland ( 2008 ) .
w. herr , lhc beam operation workshop , evian ( 2010 ) .
the cern study group on new accelerators , cern ar / int . sg/64 - 9 ( 1964 ) .
v. shiltsev , v. danilov , d. finley , and a. sery , phys . rev .
beams 2 ( 1999 ) 071001 . j .-
koutchouk , `` correction of the long - range beam beam effect in lhc using electro - magnetic lenses , '' pac01 , illinois , 2001 , pp .
16811683 ( 2004 ) .
u. dorda , f. zimmermann , w. fischer , and v. shiltsev , `` lhc beam beam compensation using wires and electron lenses , '' pac07 , albuquerque , june 2007 , pp . 15891591 ( 2007 ) .
t. sen and b. erdelyi , `` feasibility study of beam beam compensation in the tevatron with wires , '' pac05 , knoxville , 2005 , pp .
26452647 ( 2005 ) .
c. milardi , d. alesini , m.a .
pregner , p. raimondi , m. zobov , and d. shatilov , `` wire compensation of parasitic crossing in da@xmath6ne , '' epac06 , edinburgh , 2006 , pp .
28082810 ( 2006 ) .
w. fischer , o. brning , j .-
koutchouk , f. zimmermann , t. sen , v. shiltsev , k. ohmi , m. furman , y. cai , and a. chao , informal note bnl c - a / ap/291 ; icfa beam dynamics newsletter no .
220 - 225 ; proceedings of beam07 , cern-2008 - 005 , care - conf-08 - 004-hhh ( 2008 ) . j .-
koutchouk , j. wenniger , and f. zimmermann , `` experiments on lhc long - range beam beam compensation in the sps , '' proceedings epac04 , lucerne , 2004 , pp . 19361938 ( 2004 ) . f. zimmermann , j .-
koutchouk , f. roncarolo , j. wenninger , t. sen , v. shiltsev , and y. papaphilippou , `` experiments on long - range beam beam compensation and crossing schemes at the cern sps in 2004 , '' pac05 , knoxville , 2005 , pp .
686688 ( 2005 ) .
u. dorda , j .-
koutchouk , r. tomas , j. wenniger , f. zimmermann , r. calaga , and w. fischer , `` wire excitation experiments in the cern sps , '' epac08 , genoa , pp . 31763178
( 2008 ) . w. fischer , r. alforque , h.c .
hseuh , r. lambiase , c.j .
liaw , g. miglionico , t. russo , j .-
koutchouk , f. zimmermann , and t. sen , informal note bnl c - a / ap/236 ( 2006 ) .
w. fischer , r. calaga , u. dorda , j .-
koutchouk , f. zimmermann , v. ranjbar , t. sen , j. shi , j. qiang , and a. kabel , `` observation of long - range beam beam effect in rhic and plans for compensation , '' epac06 , edinburgh , 2006 , pp .
21582160 ( 2006 ) .
f. zimmermann , `` weak - strong simulation studies for the lhc long - range beam beam compensation , '' proceedings of the beam beam workshop at fermilab , fermilab - conf-01/390-t , cern lhc project report 502 ( 2001 ) w. fischer et al . , `` experiments with a dc wire in rhic , '' pac07 , albuquerque , 2007 , pp . 18591861 ( 2007 )
. t. sen and b. erdelyi , `` analytical studies of the long - range beam beam tune shifts and chromaticities , '' epac02 , paris , 2002 , pp . 12471249 ( 2004 ) .
kim and t. sen , `` simulations of beam - wire experiments at rhic , '' pac07 , albuquerque , 2007 , pp . 34923494 ( 2007 ) .
f. zimmermann , private communication .
n. abreu , `` beam beam with a few long - range encounters at long distance , '' proceedings beam07 , cern-2008 - 005 , care - conf-08 - 004-hhh ( 2008 ) .
kim , t. sen , n.p .
abreu , and w. fischer , `` simulations of wire compensator in rhic , '' hb2008 , nashville , 2008 , pp . 72 ( 2008 ) .
kim , t. sen , n.p .
abreu , and w. fischer , `` studies of wire compensation and beam beam interaction in rhic , '' epac08 , genoa , 2008 , pp .
31193121 ( 2008 ) .
kim and t. sen , phys . rev .
beams 12 , 031001 ( 2009 ) .
kim , t. sen , and w. fischer , `` long - range beam beam compensation in rhic , '' ipac10 , kyoto , may 2010 , pp .
20722074 ( 2010 ) .
a. kabel , private communication .
r. calaga , w. fischer , and g. robert - demolaize , `` rhic bblr measurements in 2009 , '' ipac10 , kyoto , may 2010 , pp . 510512 ( 2010 ) . | long - range beam beam effects are a potential limit to the lhc performance with the nominal design parameters , and certain upgrade scenarios under discussion . to mitigate long - range effects , current carrying wires parallel to the beam
were proposed and space is reserved in the lhc for such wires .
two current carrying wires were installed in rhic to study the effect of strong long - range beam beam effects in a collider , as well as test the compensation of a single long - range interaction .
the experimental data were used to benchmark simulations .
we summarize this work . |
we aim to estimate the effects of the the induced raman scattering of the strong electromagnetic wave propagating in pulsar magnetosphere .
raman scattering may be considered as a parametric decay of the initial transverse electromagnetic wave in the another electromagnetic wave and plasma wave .
the probability of this process is greatly enhanced if there many waves present in the final states .
this is induced raman scattering .
another type of induced scattering - induced brullion scattering , i.e. the decay of the initial transverse electromagnetic wave in the another electromagnetic wave and ion sound is prohibited in electron - positron plasma , since pair plasma does not support low frequency , density perturbing waves ( like ion sound wave in electron - ion plasma ) .
strong nonlinear coupling occurs when two waves beat together and the sum or difference frequency and wavelength match the frequency and wavelength of the third wave : @xmath0 in quantum language these relations may be interpreted as conservation of energy and momentum along the magnetic field , respectively .
( the transverse components of the momentum are not conserved ) .
the probability of the induced scattering depends on the number of waves in the final state , which , in turn , depends on the damping or escape rate of the daughter waves .
the energy transfer between the modes will be efficient if the energy of the pump wave is strong enough to overcome the damping losses or escape of the generated waves .
thus , the induced raman scattering is a threshold process : if the intensity of the pump wave exceeds the threshold value , the initial electromagnetic wave would start converting energy into the decay waves decreasing its amplitude exponentially .
induced raman scattering may be important in the pulsar environment in two ways .
first , it may provide en effective damping of the existing electromagnetic wave , that has been generated by some emission mechanism at the lower altitude in the pulsar magnetosphere , where the resonant conditions for the induced raman scattering were not satisfied .
this may result in a short time variability which is generally observed in pulsar radio emission .
secondly , it may provide an effective saturation mechanism for the growth of the electromagnetic wave provided that the conditions for the wave excitation by some mechanism are satisfied in the region where an effective raman scattering takes place .
the first possibility , i.e. the scattering of the existent wave , is simpler to consider . as a first approximation , we can treat the intensity of the pump wave as a constant .
then the nonlinear equations describing the wave coupling become linear in amplitudes of the decayed waves .
the exponentially growing solutions will imply an effective energy transfer from the pump wave .
as the intensities of the decay waves grow , this approximation breaks down in two cases : when the amplitudes of the decay waves become comparable to the pump wave or when the amplitude of the decay waves enter nonlinear stage and the waves start loosing energy due to some nonlinear process ( like particle trapping and acceleration ) .
the net effect of any energy loss by the decay wave is the depletion of the original pump wave . for the practical estimates
, we can neglect the nonlinear stages of evolution ( like cyclic energy transfer between the pump and the daughter waves ) and assume that if the intensity of the pump wave begins to decrease exponentially in the linear stage , then the wave decays completely . the second possibility , e.g. when the induced raman scattering provides a nonlinear saturation mechanism , is more complicated , since the intensities of all waves may be of the same order .
a considerable simplification in this case may be obtained if the damping of one of the decay modes is very strong or if it leaves the region of the resonant interaction fast enough .
then , after a short period of time ( when the intensities of the other weakly damped waves grow considerably ) the intensity of this damped mode is much smaller than the intensities of the two other modes and can be neglected . in what follows
we neglect the possible nonlinear stages of the development of langmuir turbulence ( particle trapping or quasilinear diffusion ) .
we also assume that the pump wave is broadband .
this is quite different from the conventional laboratory case of a monochromatic laser - plasma interaction .
the condition of a broadband pump wave implies that its band width @xmath1 is much larger than the typical growth rate of the decay instability @xmath2 : @xmath3 if this condition is satisfied , then we can use a random phase approximation for the statistical description of the interacting waves . on the other hand
, the condition of weak turbulence allows one to calculate the matrix elements of the interaction in the approximation of stationary phases . the general expression for the third order nonlinear current in plasma in a magnetic field has been written down by @xcite . an extremely complicated form of the corresponding expressions makes the general case of raman scattering very difficult to consider .
several important simplifications can be done when considering induced raman scattering in the pulsar magnetosphere .
first , the superstrong magnetic field allows an expansion of the currents in @xmath4 .
second , in the pair plasma with the same distributions of electrons and positrons some of the nonlinear currents cancel out since they are proportional to the third power of the electric charge ( this cancellation is exact in the unmagnetized electron - positron plasma ) .
the third , less justified approximation , is that we will make is that all the three interacting waves propagate along magnetic field .
this is an important assumption .
it allows us to simplify the consideration considerably and to obtain some analytical estimates of the effects of induced raman scattering . a short overview of the work on the nonlinear process in the pulsar magnetosphere will be appropriate here .
the possible decay processes for the transverse waves and the corresponding references are ( @xmath5 denotes transverse wave , @xmath6 denotes longitudinal wave and @xmath7 denotes a charged particle ) : ( i ) a decay of a transverse wave into two transverse wave @xmath8 ( @xcite ) , ( ii ) a decay of a transverse wave into two lagmiur waves @xmath9 ( @xcite ) , ( iii ) a decay of a transverse wave into another transverse and lagmuir wave @xmath10 ( @xcite ) ( iv ) induced scattering of transverse waves @xmath11 ( @xcite , @xcite , @xcite , @xcite ) .
the possible processes for lagmuir waves are ( i ) a decay of a lagmuir wave into another lagmuir and transverse wave @xmath12 ( @xcite ) , ( ii ) a decay of a lagmuir wave into two transverse waves @xmath13 ( @xcite ) , ( iii ) lagmuir wave merger @xmath14 ( @xcite ) , ( iii ) induced scattering of lagmuir waves ( @xcite ) .
when treating the nonlinear process involving transverse waves , the matrix elements in above works have been calculated in the drift approximation - the expansion in parameter @xmath4 was done in the very beginning ( @xmath15 is magnetic field , @xmath16 is mass of an electron and @xmath17 is the speed of light ) .
normal modes of relativistic pair plasma for the case of parallel propagation consist of two transverse waves with the dispersion relation @xmath18 and a plasma langmuir wave with a dispersion relation @xmath19 where in eq .
( [ lliafs3 ] ) the wave vectors should not be much larger than cross - over point : @xmath20 ( @xcite ) .
we have introduced here a plasma frequency in the plasma frame @xmath21 , where @xmath22 is the pulsar rotation angular velocity , @xmath23 is the multiplicity factor , @xmath24 is the lorentz factor of the plasma frame , @xmath25 is lorentz invariant temperature of plasma , @xmath26 is lorentz factor of a particle and @xmath27 is a characteristic ( dimensionless ) thermal velocity of particles , @xmath28 is ( dimensionless ) velocity .
the brackets denote averaging over the distribution function . at the point
@xmath29 the dispersion curve of plasma waves intersect that of waves in vacuum , so that for @xmath30 the plasma waves become subluminous .
the necessary requirement for that is that the distribution function falls fast enough at large values of momentum ( faster than @xmath31 ) ( @xcite ) . in the subluminous region ,
the plasma wave dispersion relation may be written as @xmath32 induced raman scattering is closely related to the induced compton scattering of waves by the plasma particles . in the conventional treatment of induced compton scattering in pulsar setting ( e.g. @xcite , @xcite ) the collective effects of the plasma are ignored . in the case of
induced raman scattering collective effects play a major role .
raman scattering is a limit of the induced compton scattering of waves by particles in a medium when the beat wave of the incoming and scattered wave becomes a normal mode of the medium ( equivalently , it falls on the `` mass surface '' of the medium dispersion equation ) . in the nonrelativistic plasma the collective treatment of the wave scattering by plasma particles
is justified if the following condition is true : @xmath40 where @xmath41 is the change in the wave number of the scattered wave and @xmath42 is the debye length . condition ( [ n3 ] ) states that the wave number of the oscillations of the electron on the beat of the two scattered waves be much less than the inverse of the debye length .
if condition ( [ n3 ] ) is not satisfied , then the beat wave will either experience a strong cherenkov damping ( if @xmath43 ) or the beat wave will not feel the presence of a medium ( if @xmath44 ) , so that the scattering process will be described as induced compton scattering .
in the relativistic plasma with the characteristic lorentz factor of the particles @xmath45 the condition ( [ n3 ] ) is modified . to find the the generalization of condition ( [ n3 ] ) to the case of relativistic plasma we note
, that condition ( [ n3 ] ) may be rewritten in therm of the phase velocities of the beat wave @xmath46 ( which is a plasma wave in case of raman scattering ) and the thermal velocities of the plasma particles @xmath47 .
condition ( [ n3 ] ) then can be rewritten @xmath48 using ( [ n5 ] ) , we find that the phase velocity of the subluminous plasma waves becomes equal to the thermal velocity of plasma particles approximately at @xmath49 .
thus , the following condition should hold for the wave - particles scattering in relativistic pair plasma to be considered as a raman scattering : @xmath50 when @xmath51 the plasma waves are strongly damped and when @xmath52 the wave - particles scattering is described as single particle compton scattering . as we will see later
( [ ttt1 ] ) ) the condition ( [ n6 ] ) is satisfied in our problem .
first we calculate the matrix elements for the nonlinear three wave interaction in the strongly magnetized electron - positron plasma when all three waves have their wave vectors along the magnetic field .
we represent the electric fields of the the three interacting waves in the form @xmath53 \label{n1}\ ] ] the indices @xmath54 refer to the scattered electromagnetic wave , plasma wave and the initial electromagnetic wave respectively , @xmath55 denotes the real part .
we consider lineally polarized waves .
for the parallel propagation the case of circularly polarized waves may be trivially reduced to the scattering of two lineally polarized waves .
implicit assumption here is that the coupling between the waves is weak , so that the waves retain their identity as eigenmodes of the medium during the times much longer than inverse of their frequencies .
we calculate transition probabilities by finding the coupling coefficients for the resonant three - wave interaction .
the interaction of two transverse and one longitudinal wave propagating along magnetic field in a relativistic plasma has been considered by @xcite . for the decay process @xmath56 and @xmath57 the coupling of the waves is described by the system of equations @xmath58 where @xmath59 and @xmath60 \mbox { } \nonumber \\ \mbox { } & & c_{\pm } = - \sum { \omega_p^ 2 q \over \omega_1 \omega_2 m } \int d { \bf p } \ , \left\ { { m\over \gamma \hat { \omega}_2 } { \partial f ( { \bf p } ) \over \partial p_z } \right . \mbox { } \nonumber \\ \mbox { } & & \mp { \partial \over \partial p_{\perp}^2 } \left [ { \omega_c p_{\perp}^2 \over \gamma^4 \hat { \omega_2 } } \left ( { k_3 \over ( \hat { \omega}_1 \mp \omega_c ) \ , ( \hat { \omega}_3 \mp \omega_c ) } - { k_1 \over ( \hat { \omega}_1 \mp \omega_c ) ( \hat { \omega}_3 \mp \omega_c ) } - { \gamma^2 v_z \over c^2 } { \hat { \omega}_1 + \hat { \omega}_3 \mp \omega_c \over \omega_1 \omega_3 } \right ) \right ] f ( { \bf p } ) \mbox { } \nonumber \\ \mbox { } & & \left . - { p_{\perp}^2 \over 2 \gamma^3 m
c^2 \hat { \omega}_2 } \left [ 1- { k_1 k_3 c^2 \over \gamma^2 ( \hat { \omega}_1 \mp \omega_c ) \ , ( \hat { \omega}_3 \mp \omega_c ) } - \right
. \right . \mbox { } \nonumber \\ \mbox { } & & \left
v_z \left ( { k_3 \over ( \hat { \omega}_3 \mp \omega_c ) } + { k_1 \over ( \hat { \omega}_1 \mp \omega_c ) } \mp { \omega_c k_1 \over ( \hat { \omega}_1 \mp \omega_c ) \ , \hat { \omega}_2 } \pm { \omega_c k_3 \over ( \hat { \omega}_3 \mp \omega_c ) \ ,
\hat { \omega}_2 } \right ) \right ] \right . \left
. \times { \partial f ( { \bf p } ) \over \partial p_z } \right\ } \label{r3}\end{aligned}\ ] ] where @xmath61 and @xmath62 is the relativistic cyclotron frequency .
we now simplify this relations for the case of one dimensional pair plasma with the same distribution functions . setting @xmath63 and introducing a positive nonrelativistic gyrofrequency @xmath64 ( @xmath65 is the sign of the charge )
we find @xmath66 the probability of the decay is ( @xcite eq .
10.124 ) @xmath67 with @xmath68 in the cold case we have @xmath69 which may be obtained from @xcite eq .
. equations ( [ r401 ] ) and ( [ r4 ] ) can be simplified in the limit of a very strong magnetic field . in the limit @xmath70
we find @xmath71 using eq .
( [ r401 ] ) the equation ( [ r5 ] ) may be rewritten in the form @xmath72 for the back scattering eq .
( [ r5 ] ) gives @xmath73 for the plasma waves on the mass surface @xmath74 and @xmath75 we can compare this probability with the unmagnetized case ( @xcite ) . in the strongly magnetized plasma the probability of raman scattering is decreased by a ratio @xmath76 .
this is similar to the suppression of the thompson scattering , whose probability is decreased by the same ratio .
it is convenient to describe the wave distribution in terms of photon occupation numbers @xmath77 where @xmath78 is the spectral energy density , which is related to the total energy density @xmath79 the kinetic equation for the raman scattering are ( @xcite , eq . ( 5.93 ) ) @xmath80 where @xmath81 are the corresponding damping or growth rates .
the most unimportant damping is the landau damping of the plasma waves @xmath82 on the particles of plasma . in ( [ kd92 ] )
the total derivative on the left hand side is @xmath83 with the second term describing the convective transport of the waves .
next we integrate over transverse wave vectors to find the system of equations describing the scattering of the waves propagating along magnetic field @xmath84 where @xmath85 and @xmath86 the two @xmath87-functions in @xmath88 allow us to perform integrations in ( [ kd919 ] ) .
@xmath89 implicit assumption in eqs .
( [ kd902 ] ) - ( [ kd9192 ] ) is that the wave vectors on the right hand side of eqs .
( [ kd902 ] ) , ( [ kd912 ] ) and ( [ kd9192 ] ) satisfy the resonance conditions .
thus in eq .
( [ kd902 ] ) , for example , the two independent variables are time @xmath5 and wave vector @xmath90 . the explicit relations between wave vectors , corresponding to eqs .
( [ kd902 ] ) , ( [ kd912 ] ) and ( [ kd9192 ] ) , are @xmath91 respectively for eqs .
( [ kd902 ] ) , ( [ kd912 ] ) and ( [ kd9192 ] ) .
the quantity @xmath92 is the cross - over wave vector , where the dispersion relation for the plasma waves intersect the vacuum dispersion relation . from eqs .
( [ ttt1 ] ) and ( [ n6 ] ) we find that the induced raman scattering occurs in the frequency region @xmath93 this frequency window ( in the plasma frame ) corresponds to the frequencies in the pulsar frame in the range @xmath94 , which is @xmath95 which falls into the observed frequency range .
in this section we consider the possible applications of the induced raman scattering to the wave propagation and spectrum formation in the pulsar magnetospheres .
there is threshold intensity of the pump wave that defines what is known as strong induced raman scattering .
the rate of induces plasmon production , which depends on the intensity of the pump wave , should exceed the rate of loss due to the various damping processes .
the end result of such strong scattering is a total depletion of the initial pump wave .
the threshold photon density for strong raman scattering is determined from ( [ kd912 ] ) @xmath96 where @xmath97 is the damping rate of plasma waves .
linear landau damping rate in the relativistic plasma ( @xmath98 ) is ( @xcite @xmath99 with @xmath100 the equations are valid for @xmath101 . with a power law distribution function @xmath102
we find @xmath103 note that as @xmath104 , i.e. when the wave vector approaches the cross - over point where the phase speed is equal to the speed of light , @xmath105 since there no particles with the speed equal to the speed of light .
the threshold spectrum of the incoming radiation is then given by @xmath106 estimating @xmath107 we find the characteristic threshold photon density : @xmath108 where @xmath109 is a magnetic field at the surface of the neutron star , @xmath110 is the multiplicity factor , i.e. the ratio of the plasma density to the goldreich - julian density @xmath111 , @xmath112 is the pulsar period , @xmath113 is the relativistic invariant temperature , @xmath114 is the distance from the neutron star , @xmath115 is the average streaming lorentz factor of the secondary plasma and we used @xmath116 .
note , that this is an _ upper _ bound on the threshold photon density ( for @xmath117 ) . at the lower frequencies @xmath118 the threshold values are much lower . for the relativistic gaussian - type distribution
@xmath119 the damping rate is @xmath120 ( compare with @xcite eq .
( 2.10 ) ) .
the spectrum of the incoming radiation is then given by @xmath121 the characteristic threshold photon density is in this case @xmath122 ( upper bound ) where we used @xmath123 .
for the comparable values of the temperature , i.e. for the similar dispersion of the energies of the plasma particles , the threshold in the case of a power law distribution is much larger than in the case of maxwell - type distribution .
this is due to the larger damping rate in the case of a power law distribution since there are more particles satisfying the cherenkov resonance in this case than in the case of maxwell - type distribution .
the estimates ( [ kd9810 ] ) and ( [ kd9811 ] ) are done for the one dimensional ( in @xmath124 ) photon density in the plasma frame
. the photon densities ( [ kd9810 ] ) and ( [ kd9811 ] ) are the one dimensional photon densities .
the energy flux per unit interval of frequencies is given in terms of photon density as follows : @xmath125 where we took into account that the dispersion relation of the electromagnetic waves is almost vacuum like , so @xmath126 .
we recall next , that both three dimensional occupation numbers @xmath127 and one dimensional photon density @xmath128 ( eq . ( refkd921 ) ) are relativistic invariants .
it follows from ( [ kd9001 ] ) that @xmath129 ( for the one dimensional flux , i.e. integrated over transverse wave vectors ) .
thus , the flux in the observer frame is @xmath130 we can estimate the observed flux ( in jankys ) at the earth : @xmath131 using eqn .
( [ kd9810 ] ) , ( [ kd9811 ] ) , ( [ kd903 ] ) and ( [ r41 ] ) we find @xmath132 for the power law and exponential distributions correspondingly .
numerical estimates give @xmath133 these are comparable with the fluxes observed from bright pulsars . since the estimated threshold intensities are upper bounds , we conclude that stimulated raman scattering may be an important factor in the formation of pulsar radio emission .
estimates for the millisecond pulsars give about an order of magnitude lower values .
cyclotron - cherenkov instability , operating in the outer parts of magnetosphere at the anomalous doppler resonance @xmath134 has been suggested as a possible mechanism of pulsar radio emission ( @xcite , @xcite , @xcite , @xcite ) . in this section
we consider the nonlinear saturation of the cyclotron - cherenkov instability by the induced raman scattering . to simplify the consideration
, we assumes that the product electromagnetic waves ( labeled by index @xmath135 ) leave the region of resonant interaction , so we can assume that their photon density is @xmath136 .
we are interested in the time asymptotic limit for the intensities of the transverse waves , excited by some instability mechanism with a growth rate @xmath137 . neglecting the wave convection
we then find from eqs .
( [ kd902 ] ) and ( [ kd9192 ] ) : @xmath138 comparing eqs .
( [ kd93 ] ) and ( [ kd9121 ] ) we find that they are identical : the threshold intensity for the strong raman scattering is equal to the saturation intensity .
this is not a surprising fact , since in both cases it is the damping of the plasma waves that controls the electromagnetic wave intensity . in section [ scateringpumpwave ]
we found that the corresponding intensities are comparable to the observed ones , so that the nonlinear saturation by the induced raman scattering may provide an observed emissivities .
we also note in this context , that at a given radius , the induced raman scattering operates in a limited frequency range ( [ n6 ] ) .
so , if the instability that produces radio emission is intrinsically broad band ( with the bandwidth of the growing mode larger than the range in which the induced raman scattering operates ) , then the effects of the induced raman scattering would produce a dip in the spectrum , corresponding to that frequency range .
if , on the other hand , the instability that produces radio emission is intrinsically narrow band with the central frequency of the instability falling into the operational range of the induced raman scattering and changing with radius to mimic the observed broadband spectrum , then the induced raman scattering can be an effective saturation mechanism for a larger range of frequencies .
in this work we considered induced raman scattering of a transverse electromagnetic wave propagating along magnetic field in a pair plasma .
we found that induced raman scattering can be an important factor for the wave propagation in the pulsar magnetosphere and as a nonlinear saturation mechanism for the electromagnetic instabilities ( cyclotron - cherenkov instability , for example ) .
the threshold intensities for strong raman scattering , which are determined by the damping rate of the langmuir waves in the relativistic plasma , are comparable to the observed fluxes from radio pulsars .
one of the major limitations of this work is the assumption of the propagation along the magnetic field .
the natural extension of this work to oblique propagation would involve several complications : ( i ) there will be a distinction in the transition probabilities between extraordinary mode ( with the electric field of the waves perpendicular to the @xmath139 plane ) and ordinary mode ( with the electric field in the @xmath139 plane ) , ( ii ) ordinary mode has an sensitive dependence on the angle of propagation , so that for the ordinary mode the raman scattering will occur only in a very limited angle around magnetic field ; the scattered waves will be a backward propagating , ( iii ) for the extraordinary mode the raman scattering is possible for large angles of propagation ; in this case both electromagnetic waves can be forward propagating .
kinematics of backwards raman scattering in the plasma frame .
the vacuum dispersion relation @xmath140 , the transverse wave dispersion relations @xmath141 and the plasma wave dispersion relations ( which starts from @xmath142 ) for the parallel propagation are shown .
the bold faced vectors represent @xmath143 .
cross - over point is given by @xmath144 and @xmath145 .
only waves with @xmath146 can be scattered . | it is shown that induced raman scattering of electromagnetic waves in the strongly magnetized electron - positron plasma of pulsar magnetosphere may be important for wave propagation and as an effective saturation mechanism for electromagnetic instabilities . the frequencies , at which strong raman scattering occurs in the outer parts of magnetosphere , fall into the observed radio band .
the typical threshold intensities for the strong raman scattering are of the order of the observed intensities , implying that pulsar magnetosphere may be optically thick to raman scattering of electromagnetic waves .
[ [ section ] ] |
the widely known class of plane - fronted gravitational waves with parallel rays ( _ pp_-waves ) has become a paradigm of exact radiative spacetimes in general relativity . found by brinkmann in 1923
@xcite and later rediscovered independently by several authors , it has been studied for decades .
the metric of vacuum _
pp_-waves can be written in the standard form @xcite @xmath0 where @xmath1 is an arbitrary function of the retarded time @xmath2 and the complex coordinate @xmath3 spanning the plane wave surfaces @xmath4 const .
when @xmath5 is linear in @xmath3 , the metric ( [ e1 ] ) represents minkowski universe since the only non - trivial components of the curvature tensor are proportional to @xmath6 . thus the simplest case for which ( [ e1 ] ) describes gravitational waves arises for @xmath7 , where an arbitrary function @xmath8 characterizes the ` profile ' of the wave .
solutions of this type are called plane waves ( or homogeneous _ pp_-waves ) and have been investigated extensively in the literature ( see @xcite,@xcite for refs ) .
this important class of exact radiative spacetimes has also been used for the construction of sandwich waves ( see e.g. @xcite-@xcite ) for which @xmath8 is non - vanishing on a finite interval of @xmath2 only .
also , the very interesting problem of the collision of two plane waves has been thoroughly studied ( see @xcite for review of the topic and refs . ) surprisingly , more general non - homogeneous _
pp_-waves with generic or sandwich wave - profiles have not been investigated as much ( a physical interpretation of some solutions of this type was proposed e.g. in @xcite,@xcite ) . on the other hand ,
much attention has been payed to impulsive waves described by the metric ( [ e1 ] ) with the function @xmath5 of the form @xmath9 , where @xmath10 is the dirac distribution .
these solutions have attracted attention recently not only for the investigation of gravitational radiation emitted during the collision of black holes @xcite,@xcite , but also in the context of field and string theories in a study of the scattering processes at extremely high planckian energies , see @xcite and elsewhere .
such impulsive waves can naturally be understood as distributional limits of suitable sequences of sandwich waves regularizing the dirac @xmath11 .
spacetimes of this type can also be obtained by boosting the schwarzschild , other kerr - newman , and dilaton black holes @xcite-@xcite , or axially symmetric solutions of the weyl class @xcite to the speed of light .
some of these solutions can be interpreted as impulsive _
pp_-waves generated by null particles of an arbitrary multipole structure @xcite .
another method for constructing general impulsive _
pp_-wave spacetimes has been proposed in @xcite : the ` scissors - and - paste ' approach is based on the removal of a null hyperplane from minkowski spacetime and re - attaching the parts by making a formal identification with a ` warp ' .
this corresponds to a specific coordinate shift @xcite .
impulsive _ pp_-waves can alternatively be introduced using a different coordinate system in which the metric is explicitly continuous @xcite , @xcite-@xcite .
it has only recently been shown @xcite that these approaches are equivalent in a rigorous sense .
they lead to identical ( unique ) particle motion , i.e. the corresponding geodesics in ( [ e1 ] ) , see @xcite-@xcite , agree with those obtained from the continuous form of the metric . in our recent works @xcite , @xcite we demonstrated that geodesic motion in the class of non - homogeneous vacuum _
pp_-wave spacetimes is chaotic .
this is the first example of chaos in exact radiative spacetimes ( chaotic behavior of geodesics in various black - hole spacetimes is well known , see e.g. @xcite-@xcite , and references therein ) . it is the purpose of the presented paper to generalize these results which remained restricted to spacetimes ( [ e1 ] ) with @xmath12 , @xmath13 , i.e. those having a _ constant _ profile , @xmath14 const . in particular , we wish to investigate the behavior of geodesics in non - homogeneous sandwich _ pp_-waves and in the corresponding impulsive limits . in section 2
we briefly summarize previous results concerning chaotic motion in _ pp_-waves with a constant profile . also , we introduce the concepts and quantities necessary for our subsequent investigation .
section 3 is devoted to geodesics in the simplest sandwich _
pp_-waves having a ` square ' profile
. a rigorous characterization of the level of the fractal structure of basin boundaries ( which separate initial conditions leading to different outcomes ) by the ` number of radial bounces ' of geodesics is introduced .
this enables us to describe quantitatively how chaos smears as the sequence of sandwich waves approaches the impulsive limit .
the same behavior of geodesics is investigated within the class of smooth sandwich waves in section 4 , where a relation between the smearing of chaotic behavior and the gradual vanishing of the outcome channels is also discussed . in section 5
explicit ( i.e. non - chaotic ) geodesics in the resulting non - homogeneous impulsive _
pp_-waves are presented and interpreted physically .
finally , in section 6 we discuss the effect of the focusing of geodesics and describe the deformation of a disc of test particles .
as shown in @xcite , @xcite the geodesic equations for ( [ e1 ] ) reduce to @xmath15\,\d \tau+\tilde v \ , \label{e4}\end{aligned}\ ] ] where @xmath16 is an affine parameter , dot denotes @xmath17 , @xmath18 are constants , and @xmath19 for timelike , null or spacelike geodesics , respectively .
it suffices to find solutions of eq .
( [ e2 ] ) . introducing real coordinates @xmath20 and @xmath21 by @xmath22 , this system follows from the hamiltonian @xmath23 where the potential is @xmath24 . for non - homogeneous _
pp_-waves given by @xmath25 the corresponding polynomial potential @xmath26 at any @xmath27 is called ` @xmath28-saddle ' .
it can by visualized in polar coordinates @xmath29 , @xmath30 where @xmath31 , in which it takes the form @xmath32 .
it was shown in a series of mathematical papers @xcite-@xcite that motion in the hamiltonian system ( [ e7 ] ) with a polynomial potential ( [ e7b ] ) where @xmath14 const .
is chaotic .
note that , interestingly , in the simplest case @xmath33 the corresponding ` monkey saddle ' potential ( after removing the factor @xmath34 ) is @xmath35 , so that we get the particular case of the famous hnon - heiles hamiltonian @xcite ( with missing quadratic terms ) which is a ` textbook ' example of a chaotic system .
this was investigated by rod @xcite who concentrated on the topology of _ bounded orbits _ in the energy manifolds @xmath36 .
the sets of orbits asymptotic to the basic unstable periodic orbits ( denoted by @xmath37 ) as @xmath38 intersect transversely .
this proves the existence of homoclinic and heteroclinic orbits and indicates the complicated structure of the flow .
these results were later refined in @xcite by showing that @xmath37 are hyperbolic , so that they admit stable and unstable asymptotic manifolds . finally , in @xcite the above hamiltonian was presented as an example of a system for which the smale horseshoe map can explicitly be embedded as a subsystem along the homoclinic and heteroclinic orbits .
it was shown in @xcite - @xcite that similar results hold for a general potential ( [ e7b ] ) with @xmath39 . therefore , geodesic motion in all non - homogeneous _
pp_-wave spacetimes with the corresponding function ( [ e7a ] ) is chaotic . in order to support these arguments for the chaotic behavior , we investigated @xcite@xcite the structure of motion by a fractal method .
complementary to the analysis described above , we concentrated on _ unbounded _ geodesics . the fractal method ( advanced in relativity in @xcite-@xcite and elsewhere )
starts with a definition of several distinct asymptotic outcomes ( given here by ` types of ends ' of all trajectories ) .
subsequently , a set of initial conditions is evolved until one of the outcome states is reached .
chaos is established if the basin boundaries which separate initial conditions leading to different outcomes are fractal .
such fractal partitions are the result of chaotic dynamics and measure an extremally sensitive dependence of the evolution on the choice of initial conditions .
we have observed exactly these structures in the system studied .
we integrated numerically the equations of motion given by ( [ e2 ] ) , ( [ e7a ] ) for @xmath39 . the initial conditions ( without loss of generality )
were chosen such that the geodesics started ( from rest ) at @xmath40 from a unit circle in the ( @xmath41)-plane .
the initial positions were parametrized by an angle @xmath42 such that @xmath43 , @xmath44 . in fig . 1 we present typical trajectories for @xmath45 .
each unbounded geodesic escapes to infinity where the curvature singularity is located _ only along _ one of the @xmath28 _ distinct outcome channels _ in the potential with the radial axis @xmath46 , @xmath47 ( in fact , it oscillates around the axis with frequency growing to infinity and amplitude approaching zero @xcite ) .
these channels represent possible outcomes of our system and we label them by the corresponding values of @xmath48 . in certain regions
the function @xmath49 representing a portrait of the basin structure depends sensitively on initial position given by @xmath30 see fig .
1 where the ( finite resolution ) results for @xmath45 are shown . in the same diagrams we plot
the function @xmath50 which takes the value of @xmath16 when the singularity is reached by a given geodesic .
the boundaries between the outcomes are fractal which we confirmed in @xcite on the enlarged detail , on the detail of the detail etc . up to the sixth level ( where numerical errors became significant ) . at
_ each level _ the structure has the property that between two connected sets of geodesics with channels @xmath51 and @xmath52 there is always a smaller set of geodesics with channel @xmath53 such that @xmath54 and @xmath55 .
this has a counterpart in @xmath50 , see fig . 1 .
the value of @xmath56 diverges on each discontinuity of @xmath49 , i.e. , on any fractal basin boundary .
there is an infinite number of peaks , each corresponding to an unstable trapped orbit which never ` decides ' on a particular outcome to infinity .
also , @xmath56 increases as one zooms into the higher levels of the fractal .
this is natural since higher levels are generated by geodesics which undergo ` more bounces ' in the inner region before escaping through one of the outcome channels .
it can immediately be observed that the above results can easily be applied to a description of geodesics in shock _
pp_-waves given by the metric ( [ e1 ] ) , ( [ e7a ] ) with @xmath57 , where @xmath58 is the heaviside step function .
it is natural to consider free test particles which are at rest ( @xmath59 ) in the flat minkowski half - space @xmath60 .
( the minkowski coordinates are given by @xmath61 , @xmath62 , @xmath63 , @xmath64 . ) at @xmath65 these particles are hit by the shock and subsequently for @xmath66 they move in the wave with a constant profile @xmath39 .
the behavior of the corresponding geodesics has been summarized in section 2 . in particular , the trajectories and the fractal structure of basin boundaries have again the form indicated in fig . 1 .
therefore , the geodesic motion in non - homogeneous shock _ pp_-waves is chaotic in a rigorous sense .
however , it could be argued that these above results concern very specific and rather ` unrealistic ' classes of _ pp_-wave solutions for which the profile function @xmath8 in ( [ e7a ] ) is constant on an infinite interval of the retarded time @xmath2 .
such waves have an ` infinite duration ' and a constant ` strength ' .
it is the purpose of the presented work to investigate more realistic non - homogeneous _
pp_-waves , namely sandwich waves of this type described by functions @xmath8 having only finite support . for an investigation of sandwich waves
it is convenient to parametrize the geodesics by the coordinate @xmath2 instead of the parameter @xmath16 ; for @xmath67 we get from eqs .
( [ e2 ] ) , ( [ e3 ] ) , ( [ e7a ] ) @xmath68 where prime denotes @xmath69 . for timelike geodesics we can simply substitute @xmath70 in the result in order to obtain the dependence on the proper time .
it is natural to start with the simplest sandwich waves having a ` square ' profile @xmath71\ , \label{e9}\ ] ] where @xmath72 and @xmath73 are positive constants .
there are flat minkowski regions in front of the wave @xmath74 and behind the wave @xmath75 . within the wavezone
@xmath76 the amplitude is constant , @xmath77 , and we can use the results described above . in particular , we can study a deformation of a ring of particles in the @xmath78-plane which are at rest in front of the wave .
the particles start moving at @xmath65 and then follow exactly the same trajectories @xmath67 as shown in fig . 1 for @xmath79 . at @xmath80
the influence of the sandwich wave ends , the potential ( [ e7b ] ) defining the outcome channels vanishes , and for @xmath81 , the particles move uniformly in different directions along straight lines in the flat space behind the wave .
consequently , the fractal structure of basin boundaries indicating chaos is ` cut ' at some level given by the value of @xmath72 : there is not enough time for particles to bounce _ arbitrarily many _
times between the potential walls . as @xmath82
, the number of bounces tends to zero so that the fractal structure is completely erased . in other words , for smaller @xmath72 , the dependence on initial conditions is less sensitive since the prediction of the outcome can be done with only a lower resolution . the narrower the sandwich wave , the ` less chaotic ' the corresponding geodesic motion .
finally , in the impulsive case given by the limit @xmath83 of ( [ e9 ] ) , i.e. @xmath84 , @xmath85 , the motion is non - chaotic .
this effect can be called a ` smearing of chaos ' in the impulsive limit of non - homogeneous sandwich _ pp_-waves .
this behavior can be described formally .
let us define the _ number of bounces _
@xmath86 of the geodesic as the number of ( local ) maxima of the function @xmath87 measuring the radial distance from the origin of the ( @xmath88)-plane .
@xmath86 represents the number of times the geodesic crosses the phase - space surface of section @xmath89 ( with @xmath90 ) before escaping to infinity .
this is a more appropriate measure here than the twist number used in literature on chaotic dynamics . as in the previous section , we consider a family of geodesics starting from rest from a unit circle @xmath91 with the angle @xmath30 parametrizing their initial positions . in fig . 2 we plot the function @xmath92 for the geodesics given by eq .
( [ e9a ] ) with @xmath93 and @xmath33 .
the sequence of graphs shows the zooming in of the fractal interval around the value @xmath94 . also , we plot the function @xmath95 which takes the value of @xmath2 when the singularity at @xmath96 is reached by a given geodesic .
it is obvious that the fractal structure described by the function @xmath92 corresponds to the structure given by @xmath95 . also , @xmath86 is a precise definition of the ` level ' of the fractal since we have demonstrated that the @xmath86-th level of the fractal structure is given by those geodesics which bounce exactly @xmath86 times before they choose one of the outcome channels .
now we can describe how the fractal structure indicating chaos arises .
let us denote by @xmath97 the set of initial conditions @xmath30 generating geodesics with @xmath98 , where @xmath99 . clearly , @xmath100 , and the complement of @xmath101 describes the fractal boundary intimately related to the fractal structure of the outcome basin boundaries ( although the discontinuities in @xmath92 and @xmath49 do not coincide ) .
now , we define a sequence of real numbers @xmath102 .
their physical meaning is the following : at @xmath103 _ all _ geodesics generated by @xmath97 have already fallen into the singularity at @xmath96 along the three channels given by @xmath104 ( after performing at most @xmath105 bounces ) . from numerical simulations visualized in fig . 2 we obtained the values @xmath106 , @xmath107 , @xmath108 , @xmath109 , @xmath110 , @xmath111 .
now , we quantitatively characterize the behavior of geodesics in the sandwich wave given by ( [ e9 ] ) .
the amplitude within the wavezone is @xmath77 but it follows from eq .
( [ e9a ] ) that we can obtain the corresponding values of @xmath95 from those for @xmath93 ( discussed above ) by a simple rescaling @xmath112 , where @xmath113 is drawn in fig . 2 .
notice that @xmath114 as @xmath115 .
it is obvious that all geodesics for which @xmath116 , i.e. @xmath117 , have time enough to fall into the singularity ( in the impulsive limit , @xmath83 , this condition can not be satisfied so that none of the geodesics considered ends in the singularity ) .
since @xmath92 is independent of the rescaling of @xmath2 , the fractal structure of initial conditions given by @xmath97 and the values of @xmath118 do not depend on @xmath73 .
consequently , the condition @xmath119 guarantees that _ all _ geodesics starting from the initial set @xmath97 will manage to fall into the singularity at @xmath96 along the three channels ( bouncing at most @xmath105-times before that ) . for a sequence of sandwich waves with @xmath120 leading to the impulsive limit as @xmath82 ,
this condition reduces to a simple relation @xmath121 which can be interpreted as follows . in order to emerge the @xmath105-th level of the fractal structure of motion the sandwich wave must have at least the duration @xmath72 such that @xmath121 . for example , if @xmath122 then all geodesics with @xmath123 bounces have time enough to choose the corresponding channel and fall along this into the singularity , i.e. @xmath124 is fully developed . for smaller @xmath72
the fractal structure is ` cut ' at lower levels @xmath97 so that the geodesic motion becomes ` less chaotic ' . for @xmath125
the fractal structure vanishes since there is no time for any geodesic ( performing at least one bounce ) to choose the outcome channel and fall into the singularity .
thus , for @xmath82 the motion is regular .
this effect occurs in all non - homogeneous sandwich _
pp_-waves with arbitrary exponent @xmath126 , only the specific values of @xmath118 for a given @xmath28 are different : for higher @xmath28 the values of @xmath118 are smaller .
we have observed that , in non - homogeneous sandwich _ pp_-waves with a short duration , many particles still remain in the inner region when the potential ( [ e7b ] ) defining the outcome channels vanishes .
these particles move through various points in different directions having different velocities when the sandwich wave ends .
subsequently , they move uniformly in the minkowski space behind the wave and it is obvious that they will _ not _ follow former channels to the singularity at @xmath96 .
in fact , any point is accessible by some trajectory .
the narrower the sandwich wave , the greater the number of particles moving ` outside ' the channels . in order to illustrate such behavior of geodesics ( and to emphasize this important aspect of the impulsive limit )
we consider another type of sandwich waves ( [ e7a ] ) with the profile function @xmath127 where @xmath128 are positive constants .
note that the corresponding radiative spacetimes are curved everywhere , becoming flat ( exponentially fast ) only asymptotically as @xmath129 . on the other hand ,
these waves are smooth since all derivatives of @xmath8 are continuous
. moreover , ( [ e10 ] ) is more appropriate for numerical integration of geodesics ( contrary to the case of simplest waves given by eq .
( [ e9 ] ) we need not apply the junction conditions which guarantee the continuity of motion at @xmath65 and @xmath80 where @xmath8 is discontinuous ) . again
, we consider a ring of free test particles which are at rest at @xmath65 ( other initial conditions , e.g. a ring which is at rest at @xmath130 lead to analogous results ) . in fig .
3 we show their geodesic motion in the wave with @xmath33 and @xmath8 of the form ( [ e10 ] ) , for different values of the parameter @xmath131 . for very small values of @xmath132 the profile remains almost constant on a large interval of @xmath2 and the trajectories ` coincide ' with those shown in fig . 1 .
the outcome channels are very well defined and the motion is chaotic . however , with a growing value of @xmath132 these channels become fuzzy . it is harder and harder to distinguish between them , so that the basin boundaries lose their fractal structure . for @xmath133 , the profile function ( [ e10 ] ) approaches the dirac delta ( in a distributional sense ) and again , in the impulsive limit the geodesic motion is not chaotic but regular .
this effect is quantitatively characterized in fig . 4 where we plot the function @xmath49 for different values of @xmath132 .
for this purpose we define three outcome windows as small intervals @xmath134 of angles around the radial axes @xmath104 of the three outcome channels ( localized sufficiently far away from the origin ) ; here we consider @xmath135 .
when a geodesic starting at @xmath30 passes through some outcome window we assign the corresponding value of @xmath136 to it .
if the trajectory does _ not _ pass through any of the three windows then we define @xmath137 which means that the geodesic does not approach infinity along the outcome channels @xmath104 . from a sequence presented in fig . 4 it is obvious that the number of geodesics localized outside the channels increases with a growing value of @xmath132 , i.e. as the sandwich profiles approach the impulsive limit . in the same diagrams we also plot the function @xmath95 .
one can observe that the number of geodesics reaching the singularity in finite values of @xmath138 rapidly decreases with a growing @xmath132 .
in particular , all geodesics outside the channel windows ( with @xmath137 ) have @xmath139 .
there is a critical value @xmath140 such that for @xmath141 _ all _ geodesics considered have @xmath142 , see fig .
these geodesics reaching @xmath96 at @xmath142 are , in fact , given by specific solutions of eq .
( [ e9a ] ) for which the condition @xmath143 is satisfied for large values of @xmath2 .
consequently , @xmath144 , @xmath145 , i.e. the acceleration is negligible and asymptotically the particles move uniformly along straight lines .
the coordinates @xmath20 and @xmath21 depend linearly on @xmath2 , so that @xmath96 is reached at @xmath146 . note that for the smooth profile ( [ e10 ] ) of the wave with @xmath33 this condition reduces to @xmath147 . once this condition is satisfied for some @xmath2 , it is valid for all greater values of @xmath2 since the left hand side grows exponentially while the right hand side only linearly as @xmath148 .
finally , we wish to find the critical value of @xmath149 . from fig . 4 it is obvious that for @xmath150 even the geodesic starting at @xmath65 from rest from the unit circle at @xmath151 ( and similarly from @xmath152 ) will reach @xmath96 with an infinite value of @xmath138 . in real coordinates , this geodesic is given by @xmath153 , @xmath154 , with @xmath155 and @xmath156 .
introducing @xmath157 we get the equation @xmath158 , @xmath159 ; the corresponding solutions for various values of @xmath131 are plotted in fig . 5 . for small @xmath132
the function @xmath160 is convex and diverges at finite values of @xmath2 . for large values of the parameter @xmath132 ,
it becomes concave so that the geodesics approaches infinity only as @xmath148 .
the boundary between these two types of behavior defines @xmath149 . from numerical simulations we obtained an approximate value @xmath161 ,
see fig .
we have demonstrated above that chaos disappears when the sequence of non - homogeneous sandwich _
pp_-waves approaches the impulsive limit . in this section
we concentrate on a description of motion in these impulsive spacetimes .
in particular , we present and discuss an analytic solution to the geodesic equations , thus proving explicitly that the motion is fully non - chaotic . for this purpose
it is convenient to use a coordinate system for impulsive _
pp_-wave spacetimes which is continuous for all values of @xmath2 @xcite , @xcite . following @xcite ,
the metric can be written in the form @xmath162\ , \d\eta|^2 - 2\,\d u\d r\ . \label{e11}\ ] ] the transformation relating ( [ e11 ] ) and ( [ e1 ] ) with @xmath9 is @xmath163 obviously , there are privileged geodesics @xmath164const . in ( [ e11 ] )
corresponding to free particles which remain at rest in flat minkowski half - space @xmath60 in front of the impulsive wave .
after the passage of the impulse these geodesics are still given by @xmath165 but the flat half - space @xmath66 behind the wave is naturally described by coordinates @xmath3 and @xmath166 ( see ( [ e1 ] ) ) in which the motion is uniform . from eq .
( [ e12 ] ) it follows that , for impulsive waves given by @xmath167 ( so that @xmath168 ) , the motion in the transversal plane is described simply by @xmath169 .
if we parametrize the initial position of each particle by @xmath170 , the motion behind the impulse ( @xmath66 ) is explicitly given by @xmath171\,u \
, \nonumber\\ & & y(u ) = \rho_0\sin\phi_0\ + c\rho_0^{n-1}\sin[(n-1)\phi_0]\,u \ , \label{e14}\end{aligned}\ ] ] so that @xmath172 the motion is uniform , i.e. the trajectories are straight lines with the velocity of each particle @xmath173,\sin[(n-1)\phi_0])$ ] being constant .
thus , the inclination of the straight trajectory in the transversal @xmath78-plane is @xmath174 , and the speed is @xmath175 .
notice that the speed depends on @xmath176 while the direction of motion on @xmath177 only
. the above geodesics can easily be visualized .
we do not present their trajectories which would be very similar to those in sufficiently narrow sandwich waves , such as the @xmath178 , @xmath33 case shown in fig .
3 . instead
, we draw in fig .
6 ` complementary ' pictures showing a deformation of a ring of free test particles ( for @xmath179 , @xmath33 ) . the initial circle at @xmath65 is continuously deformed into smooth curves with @xmath28 growing loops which are visualized here as sequences of 4 , 8 , 13 and 22 consecutive steps @xmath180 , i.e. the largest connected curve describes the deformation of the ring at @xmath181 . similarly , in fig .
7 the deformation is shown for @xmath182 and @xmath183 .
the most distant particles are those which started at @xmath184 , @xmath185 with @xmath186 so that @xmath187 ; they exactly follow the radial axes of the outcome channels which would be present for corresponding
_ pp_-waves with a constant profile . on the other hand , for @xmath177 given by @xmath188 , we get @xmath189 . it can be observed that for small values of @xmath2 the deformation of the ring agrees with that shown in fig 8 . of ref .
@xcite describing chaotic motion in the non - homogeneous _
pp_-waves with a constant profile .
the principal difference occurs for large @xmath2 . in the impulsive case
there are no subsequent loops arising from more and more bounces in the inner region so that the circle can not be deformed in a fractal way with different segments moving to different outcome channels . instead , there are no channels to the singularity
, the trajectory of each particle is explicitly given by ( [ e14 ] ) , and the motion is non - chaotic .
these results can easily be generalized to the case when the test particles are not at rest initially . from the metric ( [ e1 ] )
it is obvious that , using the coordinate @xmath3 , the motion must be uniform in both flat half - spaces , i.e. @xmath190 , @xmath191 , where @xmath192 are constants .
the relation between these parameters can again be found using the continuous form of the impulsive metric ( [ e11 ] ) . by solving the corresponding geodesic equations
we could obtain @xmath193 .
although this function is very complicated for @xmath66 , it has the property that @xmath193 and @xmath194 are continuous ( even at @xmath65 ) . using this fact and eq .
( [ e12 ] ) , which for @xmath167 reduces to @xmath195 , we obtain @xmath196 and @xmath197 , i.e. @xmath198\,u+\eta_0\ .
\label{e15}\ ] ] therefore , the trajectory of each particle in the transversal plane is a continuous but refracted straight line with a discontinuity in the velocity at @xmath65 given by @xmath199 .
notice that the value of the jump depends on the position @xmath200 of the particle at @xmath65 only , not on its actual velocity .
by parametrizing @xmath201 we immediately obtain the discontinuity in velocity , @xmath202,\sin[(n-1)\phi_0])$ ] .
thus , the result is trivial : if the particle is not at rest initially , behind the wave its constant velocity merely superimposes to the effect of a characteristic jump in velocity given by the impulse . by denoting @xmath203 , @xmath204 , and similarly @xmath205 , @xmath206 ,
we can rewrite the expression for @xmath207 as @xmath208 $ ] and @xmath209)$ ] , which generalize to non - axisymmetric cases the ` refraction formula ' for deflection on ( null ) geodesics in the axisymmetric aichelburg & sexl spacetime @xcite ( see e.g. @xcite , @xcite , @xcite ) . so far , we have concentrated on a description of motion in the transversal plane .
we should also comment on behavior in the longitudinal direction . from the form of the metric ( [ e11 ] ) it follows that , for geodesics @xmath165 , the coordinate @xmath210 is given by @xmath211 , @xmath212 being constants .
moreover , @xmath210 is continuous for all @xmath2 . due to the eq .
( [ e13 ] ) , there is a jump in @xmath213 at @xmath65 given by @xmath214 depending both on @xmath176 and @xmath177 .
since the minkowski coordinates are @xmath215 $ ] , @xmath216 $ ] , there is a discontinuity @xmath217 at @xmath65 .
this ` shift ' effect of impulsive wave on geodesics is well - known for the axisymmetric aichelburg & sexl spacetime @xcite , see @xcite , @xcite , @xcite , @xcite and elsewhere . a solution of the geodesic ( and geodesic deviation ) equations in general impulsive _
pp_-wave spacetime has been presented in @xcite-@xcite .
starting from the distributional form of the metric ( [ e1 ] ) , one has to deal with ill - defined products of distributions . in order to obtain correct results in a mathematically rigorous fashion
, careful regularization procedures and delicate manipulations during calculation of the distributional limit are required .
in fact , the colombeau theory of generalized functions providing a suitable consistent framework has to be applied . using this rigorous solution concept ,
it was shown in @xcite that in the impulsive limit the geodesics are totally independent of the regularization , i.e. on the particular shape of the sandwich wave ( the impulsive wave ` totally forgets its seed ' ) .
here we obtained an identical explicit form of geodesics starting from the continuous form of the impulsive metric ( [ e11 ] ) .
it was demonstrated in @xcite that these two approaches are equivalent in a mathematical sense ( even if the transformation ( [ e12 ] ) , ( [ e13 ] ) is discontinuous ) .
however , our main goal here was not to re - derive the known geodesics but to concentrate on their physical description and visualization .
obviously , geodesics with parallel trajectories in flat half - space in front of a sandwich or impulse wave are refracted .
a natural question arises whether a specific character of the refraction leads to some form of focusing of the corresponding geodesics .
such an astigmatic focusing effect is well - known for plane waves , and results in interesting caustic properties thoroughly investigated in @xcite . for principal reasons ( non - existence of a diagonal rosen form of the metric , see eq .
( 15 ) in @xcite , or the presence of chaos in geodesic motion ) it would be a very difficult task to reproduce these results for non - homogeneous sandwich waves . therefore , we restrict ourselves to impulsive gravitational waves only
. it immediately follows from eqs .
( [ e14 ] ) , ( [ e16 ] ) that particles staying at fixed @xmath218 in front of the impulse , move at @xmath66 according to @xmath219=c^2\rho_0^{2n-2}u -\textstyle{\frac{2}{n}}c\rho_0^n\cos(n\phi_0)\ , \label{e18}\end{aligned}\ ] ] where @xmath220 are minkowski coordinates behind the impulse , @xmath221 and @xmath222 are given by ( [ e14 ] ) .
in particular , for impulsive plane waves ( @xmath223 ) we get @xmath224 , @xmath225 and @xmath226=c(1+cu)y_0 ^ 2-c(1-cu)x_0 ^ 2 $ ] , where @xmath227 , @xmath228 . consequently , at @xmath229 one
gets @xmath230 , @xmath231 and @xmath232=cy_0 ^ 2 $ ] , so that all particles meet at @xmath233 , independently of their initial position @xmath234 in the @xmath235-plane . the generically quadratic surface ( hyperbolic paraboloid ) @xmath236const .
degenerate at @xmath237 to a parabolic caustic line @xmath233 , @xmath238 .
this is shown in fig . 8 which visualizes the coordinate singularity formation .
it plays a crucial role in colliding plane - wave solutions where the corresponding global ` fold ' singularities arise , as described in @xcite .
consequently , these spacetimes are globally hyperbolic in contrast to single plane - wave solutions where the focusing effect on null cones forbids an existence of a spacelike hypersurface which would be adequate for the global specification of cauchy data @xcite . here
we wish , however , to investigate key features of the focusing effect in more general non - homogeneous impulsive _
pp_-wave spacetimes with @xmath239 . from the deformation of a ring of free particles indicated in fig . 7
it follows that _ only @xmath28 privileged particles can collide _ at the origin ( in contrast to the plane - wave case where particles with arbitrary @xmath234 are focused along the @xmath233 line ) .
it can easily be derived from eq .
( [ e14a ] ) by setting @xmath240 that only those @xmath28 particles with @xmath188 reach the origin of the transversal ( @xmath88)-plane , simultaneously at @xmath241 .
notice that particles starting from a larger circle focus sooner .
this is understandable : they have to travel a bigger distance @xmath176 , however , their speed is also bigger , @xmath242 , so that the travel time is indeed @xmath243 .
note that the occurence of such focusing is associated with the coordinate singularity in the rosen form of the metric ( [ e11 ] ) . at @xmath244
the circle is deformed into a curve with @xmath28 cusps , see fig .
7 . the value of @xmath245 can be found from the condition @xmath246 , where @xmath247 $ ] , with @xmath248 being given by ( [ e14 ] ) .
this leads to equation @xmath249 admitting a unique positive solution @xmath250 = u_f/(n-1)$ ] .
for @xmath179 we get a simple expression @xmath251 which is in agreement with fig .
in fact , the shape of deformation of the initial circle @xmath252 of test particles in the transversal plane has a beautiful geometrical interpretation : at any time @xmath2 , _ the circle is deformed into a curve called a hypotrochoid_. such a curve is generated by a point @xmath253 attached to a small circle of radius @xmath254 rolling around the inside of a large fixed circle of radius @xmath255 , with @xmath256 being the distance from @xmath253 to the centre of the rolling circle @xcite .
( [ e14 ] ) for particle motion behind the impulsive wave are just the parametric equations for a hypotrochoid with the identification @xmath257 where @xmath177 is the parameter .
we observe that @xmath258 so that _ the hypotrochoids are closed curves with @xmath28 loops _ in our case .
also , the parameter @xmath256 grows linearly with @xmath2 . at @xmath65
we have @xmath259 and the curve is just the initial circle of radius @xmath260 . at @xmath244
the parameter has the value @xmath261 and the hypotrochoid reduces to the curve called a _ hypocycloid _ which has @xmath28 cusps .
note that it degenerates for @xmath223 to a line segment , a 3-cusped hypocycloid is called a deltoid , a 4-cusped hypocycloid is called an astroid ( see fig . 7 for @xmath33 and @xmath262 case ) . finally , at the focusing time @xmath263 we get @xmath264 so that the hypotrochoid @xmath28-times intersects the origin .
the corresponding curve is called a _ rose _ since it resembles a flower with @xmath28 petals ( for @xmath223 the rose degenerates to a line since @xmath265 ) .
it should be emphasized that the above behavior of geodesics in the @xmath78-plane describes only part of the overall deformation .
we have to consider not only the transversal deformation but also the motion in the @xmath266-direction described by ( [ e18 ] ) , as we have done for the plane - wave case in fig .
8 . at @xmath65 the longitudinal deformation
@xmath267 $ ] caused by the impulse is given by the second term on the right hand side of eq .
( [ e18 ] ) , which is exactly the shift @xmath214 .
for @xmath66 the first term increases with @xmath2 ; at @xmath268 the deformation of the whole disc @xmath269 , @xmath270const .
, becomes non - negative and subsequently grows to positive values , uniformly with @xmath2 . in combination with the deformation in the transversal plane we get an interesting new effect of ` wrapping up ' of the inner part of the disc into a closed caustic surface as shown in fig .
note that this behavior is specific for non - homogeneous _
pp_-waves and does not occur in impulsive plane waves since @xmath271 when @xmath223 .
we have shown for non - homogeneous _ pp_-wave spacetimes that the motion of free test particles is chaotic . however , for non - homogeneous sandwich gravitational waves the chaotic behavior smears as the duration of the wave tends to zero . in the limit of impulsive waves
the motion is regular . the focusing effect of non - homogeneous impulsive waves
can also be described explicitly and leads to a formation of a specific closed caustic surface .
it would be an interesting task to investigate from this point of view geodesics in other classes of exact radiative spacetimes and elucidate an open question whether such a type of behavior is common in general relativity . also , it is obvious that sandwich _
pp_-wave solutions may serve as local models of gravitational waves far away from a radiating source .
in realistic situations these waves should contain non - homogeneous components whose influence on test particles has been described above . in particular ,
chaotic - type effects on null geodesics could in principle be observable after the passage of the sandwich wave as characteristic `` chaotic deformation patterns '' of an observer s view of the sky , i.e. as peculiar changes of positions of stars and galaxies .
these specific astronomical consequences will be investigated elsewhere .
we acknowledge the support of grant no gacr-202/99/0261 from the czech republic .
we thank the developers of the software system famulus ( tom ledvinka in particular ) which we used for computation and drawing of all the pictures .
also , we thank jerry griffiths for all his help with the manuscript . *
fig.1 * geodesics starting from a unit circle escape to infinity only along one of the @xmath28 channels ( left ) .
the functions @xmath49 and @xmath50 indicate that basin boundaries separating different outcomes are fractal ( right ) .
* fig.2 * plots of the functions @xmath92 and @xmath95 in the sequence of ever narrower intervals exhibit their fractal structure .
obviously , the number @xmath86 of bounces of each geodesic is an appropriate measure of the fractal level .
* fig.3 * geodesics in the smooth sandwich gravitational waves given by eq . ( 10 ) , @xmath33 . for small values of the parameter @xmath131 , most of them escape to infinity only along the well - defined three channels . for larger values of @xmath132
these channels become fuzzy and the basin boundaries separating different outcomes are losing their fractal structure . in the impulsive limit ( @xmath133 ) the motion is regular .
* fig.4 * plots of the functions @xmath49 and @xmath95 describing behavior of geodesics in smooth sandwich gravitational waves .
when the parameter @xmath131 is small , most geodesics reach the singularity in finite value of @xmath138 . for @xmath272
all geodesics have @xmath142 .
also , for higher @xmath132 the number of geodesics outside the three outcome windows ( @xmath273 ) is greater . * fig.6 * the deformation of a ring of free test particles which are at rest initially in the tranvsveral @xmath78-plane by the influence of an impulsive gravitational wave with @xmath276 at @xmath277 and @xmath278 with steps @xmath180 .
bellow is the 3d visualization of the deformation @xmath279 .
* fig.7 * deformation of a unit ring of particles by impulsive waves with @xmath84 at @xmath183 for @xmath182 . at @xmath280
all the privileged @xmath28 particles focus at the origin of the transversal plane .
the curves are hypotrochoids , as explained further in the text .
* fig.8 * deformation of a disc consisting of free particles @xmath281 , @xmath282 . by an impulsive plane wave @xmath283
the dics is deformed into a hyperbolic paraboloidal surface at @xmath236const.@xmath284 in the minkowski space behind the wave . at @xmath285
the surface degenerates to a caustic parabolic line .
* fig.9 * deformation of a disc @xmath281 , @xmath282 of particles by an impulsive non - homogeneous wave @xmath276 at @xmath286 and @xmath287 in the minkowski space behind the wave . at @xmath285
all the privileged three particles collide and with the deformation in the longitudinal @xmath266-direction a closed caustic surface is created . | recent results demonstrating the chaotic behavior of geodesics in non - homogeneous vacuum _ pp_-wave solutions are generalized . here
we concentrate on motion in non - homogeneous sandwich _ pp_-waves and show that chaos smears as the duration of these gravitational waves is reduced . as the number of radial bounces of any geodesic decreases , the outcome channels to infinity become fuzzy , and thus the fractal structure of the initial conditions characterizing chaos is cut at lower and lower levels . in the limit of impulsive waves ,
the motion is fully non - chaotic .
this is proved by presenting the geodesics in a simple explicit form which permits a physical interpretation , and demonstrates the focusing effect .
it is shown that a circle of test particles is deformed by the impulse into a family of closed hypotrochoidal curves in the transversal plane .
these are deformed in the longitudinal direction in such a way that a specific closed caustic surface is formed .
= -1.3 cm = -2.5 cm pacs numbers : 04.20.jb ; 04.30.-w ; 05.45.+b ; 95.10.fh keywords : _ pp_-waves , chaotic motion , impulsive gravitational waves |
where the proton ( or generally nucleon ) mass comes from remains , even after the discovery of higgs boson , as one of the great mysteries of nature . in this note
i explore the possibility that the answer could perhaps be found in nuclear physics .
given that the mass of the proton is very accurately measured , @xmath3 mev , more than 99.9% of the visible " mass around us all the way down to atoms can be accounted for in great accuracy by adding the number of nucleons involved in the system .
even that of the nucleus which is in the core of atoms is accounted for up to 98% .
the strong interactions taking place inside the nucleus are now quantitatively described by quantum chromodynamics ( qcd ) .
so the mass of a nucleus is nearly completely given by the sum of the mass of the nucleons in the nucleus with only a small correction of binding energy , less than 1% of the proton mass , and even that small value can be fairly well explained , though in a highly intricate way , by qcd .
this additive accountability , however , ends abruptly and singularly at this point .
the proton mass , the source for the nuclear mass , is no longer accounted for as a sum of something . " in qcd , the relevant constituents are quarks , and for the nucleon the ( current ) quarks involved , up " and down , " are very light , less than 5 mev .
thus how the proton mass arises from its constituents must be drastically different from how the mass of a nucleus comes about .
effectively , qcd for the proton is a theory for mass without masses"@xcite . for most of particle physicists ,
however , the proton mass is satisfactorily explained .
" hence the end of the story ? in wilczek s words@xcite , here there are no uncontrolled approximations , no perturbative theory , no cutoff , and no room for fudge factors . a handful of parameters , inserted into a highly constrained theory of extraordinary symmetry , either will or wo nt account for the incredible wealth of measured phenomena in the strong interactions . and
it certainly appears that they do . "
indeed lattice simulation with qcd with no mass terms and with the heavy quarks @xmath4 , @xmath5 and @xmath6 ignored , termed qcdlite " by wilczek , predicts @xmath7 of the proton mass .
if pressed to explain in more detail , they will then say it comes from the massless gluons interacting with the nearly massless , high - momentum quarks , winding up confined in a bag . " in a nutshell , one may say the mass is generated due to the breaking of chiral symmetry associated with the nearly massless quarks by the confinement@xcite . in the standard paradigm , the chiral symmetry spontaneously broken leading to the mass generation
is characterized by the quark condensate , @xmath8 , the order parameter for chiral symmetry .
to nuclear physicists , ironically , this explanation raises more questions than answers .
this may be due to the fact that confinement is very poorly if at all understood .
for mathematicians , it is one - million dollar clay millenium problem , " still to be solved .
it may even be , as some argue , that deconfinement , " corollary to confinement , does not take place within the framework of qcd@xcite . for nuclear physicists ,
it presents numerous puzzles , some of which bear directly on observables as i will describe below .
one extremely simple picture of how confinement could lead to the spontaneous breaking of chiral symmetry is to look at a massless dirac particle let me call it quark swimming in one space dimension to the right on top of the dirac sea .
suppose there is an infinitely tall impenetrable wall on the right of what i will call the jain cell . "
the swimmer , bumping into the wall , has no choice but to turn back or drown .
now chiral symmetry forbids the swimmer from turning and swimming back unscathed to the left on _ top _ of the dirac sea .
it could swim back on top of the sea only if the quark picked up a mass , that is , if chiral symmetry were broken
explicitly by the wall .
the wall of the jail - cell " could then be the source of the mass@xcite .
this is the well - known mit bag picture of the hadron . but how does this give the mass ?
the jail wall , one might say , is erected by the as - yet un - understood strong quark - gluon dynamics and the existence of the mass is then the consequence , in the qcdlite , of the chiral symmetry broken by the jail wall , making the quark condensate nonzero .
this may sound fine , but there is something missing here , and that is the nambu - goldstone theorem that chiral symmetry realized with a non - zero condensate in massless theory like qcdlite must have zero - mass pions , namely , nambu - goldstone ( ng ) bosons .
it has been understood since many years that pions are absolutely essential for nuclear physics , although it was not clear how they figured precisely until meson - exchange currents were understood with the advent of chiral perturbation theory .
there are many different ways to bring in pions and have them couple to the quarks in the nucleon .
this accounts for the variety of models found in the literature that incorporate chiral symmetry , such as chiral bag , chiral soliton , cloudy bag etc .
it is very possible that different ways , constructed in consistency with symmetries and invariances involved , could lead to qualitatively similar results . here
i will adopt the approach called chiral bag " since it is in that picture that i can address the issues involved with some confidence . in the chiral bag model@xcite ,
both the pions and the confinement of quarks / gluons are taken to be indispensable in the structure of the elementary proton as well as that of the nucleus . in this model ,
the pion , a nb boson ( in qcdlite ) , is not just a propagating field but contains highly nonlinear mean - field components that encode topology .
let us then look again , this time taking into account the intricacy involving the ng boson , at the one - dimensional swimmer confined in the jail - cell@xcite .
as the quark @xmath9 bumps into the wall , it is subject to a boundary condition imposed by the wall , @xmath10 where @xmath11 is the outwardly - directed unit 2-vector and @xmath9 is the quark field .
this is a classical condition imposed by the vector current conservation .
one can easily see that @xmath12 , so the quark charge is confined .
but we are forgetting here that there is pion that couples , due to chiral invariance , to the quark at the surface .
the coupling modifies the boundary condition to in^_=-e^i_5/fwhere @xmath13 is the one - dimensional
pion " field and @xmath14 is a constant one dimensional analog to the pion decay constant @xmath15 .
classically the charge is still conserved .
but quantum mechanically , the coupling brings in divergences , and as is well known in field theory the divergence has to be regulated to make the theory sensible .
the regularization turns out make the fermion ( quark ) charge @xmath16 no longer conserved , making it leak out of the bag as @xmath17 .
formally this is a consequence of the vector - current anomaly generated at the bag boundary when the axial current conservation is imposed .
for the quark swimmer , what happens at the bag wall then is that the helicity conservation forces the swimmer to plunge into the dirac sea and swim back to the left under the dirac surface .
this plunge into the occupied dirac sea , made possible due to what is called infinite - hotel phenomenon"@xcite , however , makes the fermion charge disappear from the system .
but we have a theory in which the charge should be conserved .
so what happens to the fermion charge that disappears as the quark plunges into the sea
? here the penny drops !
the leaking charge is picked up by the soliton from the pion outside of the bag . in terms of the so - called chiral angle " @xmath18 where @xmath19 is the position of the bag wall , the leaking charge depends on the size of the cell , @xmath20 .
the soliton carries away precisely this leaked charge . for what s referred to as the magic angle " @xmath21 ,
the fermion charge is precisely 1/2 inside and 1/2 outside of the bag . for any angle @xmath22
, it is partitioned in such a way that the total is precisely 1 .
involving topology , exactly the same leakage of charge should take place also in ( 3 + 1 ) dimensions and indeed it does .
the soliton that carries the leaking charge is the famous skyrmion@xcite .
this was first shown in @xcite for the case of the magic angle " in @xmath23 dimensions where the baryon charge is partitioned half - and - half as in @xmath24 dimensions .
it turns out , however , that even though many nuclear properties seem to work well when described at the magic angle , " there is nothing special except for numerical efficiency about this angle .
the partition does indeed take place for _ any _ chiral angle as was shown quite convincingly by goldstone and jaffe@xcite . in this picture
, the fermion ( baryon ) charge is _ not confined _ inside the bag .
this phenomenon is dubbed
cheshire cat phenomenon"@xcite drawing an analogy from the famous smile of cheshire cat in
alice in the wonderland " of lewis carol .
one might attribute the above cc phenomenon to the topological quantity .
what about processes that are not topological ? what about , for instance , the color charge that has to do with confinement but is not connected to topology ?
does the color charge leak ?
the answer is : yes it also leaks .
this is because just like the baryon charge , the color charge is also broken by the vector anomaly .
classically the color is confined within the bag by the vector current conservation , but quantum mechanically it leaks out at the bag boundary . here
the culprit is the @xmath25 associated with the @xmath26 anomaly .
lodged outside of the bag , like pions , the @xmath25 couples to the quark at the boundary and induces a color electric field normal to the surface .
but unlike the baryon charge which can be transferred to the soliton in the pion outside , there is nothing outside that can carry the leaking color charge .
the @xmath25 causing for the anomaly is color singlet .
the only way color gauge symmetry can be preserved is then to append to to the wall a color - symmetry - breaking counter term to stop the flow of quantum - induced color@xcite .
note that this counter term is classical , so the lagrangian at the classical level is color - gauge non - invariant .
this is contrary to what usually happens in gauge theories without boundaries where symmetries conserved in the lagrangian are broken quantum mechanically , such as , e.g. , @xmath26 anomaly , trace anomaly etc .. ( or matter " ) contribution ( @xmath27 ) , ( b ) the contribution of the static gluons due to quark source ( @xmath28 ) , ( c ) the gluon casimir contribution ( @xmath29 ) , and ( d ) the sum of ( a)-(c ) ( @xmath30 . the shaded area corresponds to the range admitted by experiments.,width=377 ] how does the cheshire cat phenomenon manifest itself in physical processes that involve neither baryon charge ( topology ) nor color charge ( gauge invariance ) directly ?
a good example for answering this question is the flavor singlet axial coupling constant ( fsac ) @xmath31 in the current @xmath32 .
the fsac became famous because of the so - called proton spin crisis " which turned out be a false alarm .
the @xmath31 in general is a form factor depending on the momentum transfer involved which in some kinematical condition is related to the proton spin .
but the axial charge at zero momentum transfer that i will denote as @xmath33 has nothing directly to do with the proton spin . what is of relevance to what i am discussing here intricately involves quarks and gluons inside the bag and @xmath34 and @xmath25 outside with the baryon charge fractionization and the color - charge anomaly cancelation
so in the chiral bag model topology as well as confinement figure in @xmath35 .
one can write down a simple chiral - bag lagrangian that encodes all the above properties .
one was written down in @xcite and applied to the explicit calculation of the fsac in @xcite .
the details are given in the references , so i wo nt go into them .
the result , shown in fig
. [ fsac ] , illustrates clearly how the cheshire cat principle works .
while the individual contributions from inside and outside are widely varying in terms of the confinement radius " @xmath19 the total has , within the approximations made , no visible dependence on @xmath19 . the bag model ( e.g. ,
mit bag " ) and the skyrmion are in a sense equivalent .
this is a signal that at low energy / momentum , physical properties may very well be blind " to the confinement " size .
in fact , one can even formulate , within the framework of chiral bag , the notion that the bag size " is a gauge artifact , meaning that one can pick any @xmath19 without changing physics@xcite . in this description
, deconfinement seems irrelevant .
so what can one say about the source for the proton mass based on the cheshire cat phenomenon ?
up to here , it only connects the skyrmion to the bag but says nothing on the mass .
we will come to this matter below .
according to a general argument for baryon structure in qcd deduced in the large @xmath36 limit , the only nonpertubative tool available for approaching qcd in nonperturbative regime of high density , the constituent quark model should be equivalent to the skyrmion model .
i will exploit this equivalence in what follows . in nambu s pathbreaking paper@xcite ,
it is stated in conclusion ... our model hamiltonian , though very simple , has been found to produce results which strongly simulate the general characteristics of real nucleons and mesons .
it is quite appealing that both the nucleon mass and the pseudo - scalar ` pion ' are of the same dynamical origin , and the reason behind this can be easily understood in terms of ( 1 ) classical concepts such as attraction or repulsion between particles , and ( 2 ) the @xmath37 symmetry . " for convenience i will refer to this statement as defining nambu mechanism . " in the context of modern development , it seems more appropriate to replace the proton " in the nambu scenario by constituent quark " that i shall equivalently refer to as quasiquark " ( denoted @xmath16 ) and to view the proton ( nucleon ) as a bound state of three @xmath16s with the mass @xmath38 generated by the nambu mechanism . in this model a meson ,
say , @xmath0 meson , is a constituent - quark - constituent - antiquark ( @xmath39 ) bound state .
the interactions between @xmath16s are supposed to be weak , thereby justifying the term quasiquark " in analogy to quasiparticles in many - nucleon systems .
this @xmath16 picture is justifiable in the large @xmath36 limit from qcd@xcite .
in fact , this description is known to work very well , encompassing all light - quark hadrons , mesons and baryons . if the quasiquark mass @xmath40 arises from the nambu mechanism of chiral symmetry breaking ( ccb ) , then if one were to unbreak " the symmetry by density ( or temperature ) such that @xmath41 , then we should expect that @xmath42 . 0.3 cm @xmath43 * q(uasiquark ) scenario * : in approaching @xmath44 at which @xmath41 , the masses would scale @xmath45 and @xmath46 in such a way that @xmath47 . here
it is assumed that the quasiparticle picture holds in dense matter .
the sharp bag boundary set up in the discussions given above could very well be an artifact of approximation . in nature
, the leakage process could be smooth .
the cheshire cat implies that when the bag is shrunk , all properties of quarks and gluons will be transferred to the topological soliton , the skyrmion .
the mass generated in the qcdlite by confinement must then be lodged in the lump of soliton . in this picture
there is nothing special with single skyrmion , proton .
it can be a nucleus with the mass number @xmath48 given by the winding number @xmath48 . in the mathematics community , this feature is being exploited to unify in terms of complex geometry the proton structure to the nuclear as well as atomic structure@xcite .
seen from this vista , there is nothing special about the proton from , say , the carbon-12 nucleus .
indeed , the nucleus @xmath49c including the subtle structure of hoyle states can be understood as a skyrmion with winding number 12@xcite .
it is equally well described in chiral perturbation theory with baryonic chiral lagrangians@xcite and also in no - core shell model@xcite .
the dynamics involved in these descriptions , ranging from skyrmions to effective field theories la weinberg s folk theorem to the traditional shell - model implemented with symmetry properties of chiral symmetry seem to tell more or less the same story .
the key question is : how can one decode the source of the proton mass " going from skyrmions to quark - gluon structure as indicated by the cheshire cat phenomenon which is most likely different form the @xmath16 scenario ? to address this question , let us unbreak " the symmetries involved
e.g. , uncondense " the quark condensate by going to highly dense baryonic matter as one expects to encounter in compact stars .
to do this , i consider putting skyrmions on crystal lattice and squeezing the lattice size . if one accepts that ( 1 ) the skyrmion structure is a good description of nucleons in the large @xmath36 limit and ( 2 ) skyrmions on crystals are a reliable model for baryonic matter at high density , then the first observation is the remarkable phenomenon that as density is increased beyond the normal nuclear matter density @xmath50 @xmath51 , the skyrmions in baryonic matter fractionize into half - skyrmions . at present , due to incomplete understanding of the model , it is not feasible to give an accurate value for the transition density , but based on certain sophisticated lagrangians to be described below combined with available phenomenology , the transition density is estimated to be @xmath52 .
it can not be much lower as there is no indication from available experimental data .
it can not be much higher , for in that case , the model can not be trusted .
this is just about the density regime that one expects to be able to probe in various accelerators for nuclear physics , for example , raon in korea . also that a soliton can turn into various fractionized solitons is widely conjectured and in some cases observed in nature (
see for instance @xcite ) . for the skyrmion in question
, one can actually mathematically verify that the fractionization of a skyrmion takes place in the presence of a certain potential term of heisenberg type"@xcite .
such a potential is not present in ( generalized ) lagrangians currently studied so it is somewhat academic , but as i will describe there is scale symmetry that could generate such a potential in dense medium@xcite . in any event
, there is an indirect indication for half - skyrmion structure already in the @xmath53 particle in which four nucleons are strongly bound@xcite .
the fractionization on crystal lattice at a density @xmath54 has been observed since some time in the skyrme model ( with pions only ) or generalized models with massive degrees of freedom @xcite . as indicated in the mathematical investigation@xcite ,
the fractionized skyrmions are most likely confined , non - propagating degrees of freedom . of great significance to what i am discussing here
is that when the skyrmions fractionize into half - skyrmions , the quark condensate @xmath55 , which is non - zero both globally and locally in the skyrmion phase , vanishes when averaged over the crystal lattice .
let me denote this averaged condensate by @xmath56 .
while vanishing on the average , however , the chiral condensate @xmath55 supports chiral density wave in the half - skyrmion phase@xcite .
this means that there is a sort of phase transition in the change of topology with the condensate vanishing on averaging .
but there is still pion in the matter , with nonzero @xmath15 , hence chiral symmetry is still broken .
note that there is no obvious local - field order parameter .
it is an unusual state of matter , but such states are observed in condensed matter physics@xcite .
the questions then are : is this phenomenon , if not an artifact of crystal structure , relevant to nuclear physics ? if it is , how to implement this phenomenon , atopology change , " in the proton mass problem ?
the answer to the first question is yes and the answer to the second is what i turn to next .
the first question raised above is generic and does not depend on what other degrees of freedom than pion are involved .
however the answer to the second question involves both vector mesons and scalar mesons .
they influence neither the topological structure of matter which is controlled by the pion field only nor the cheshire cat phenomenon .
but it is found that they play extremely important roles for nuclear interactions within the framework of the effective field theory model i am using .
two hidden symmetries are found to be involved : one , hidden gauge symmetry for the vector mesons @xmath57 and the other , hidden scale symmetry for a low - mass scalar meson ( @xmath58 ) . from here on i will focus on the @xmath0 meson ( the hidden symmetry argument applies also to @xmath59 , though in a different way in dense medium ) and the dilaton @xmath58 associated with the spontaneous breaking of scale symmetry .
nuclear interactions that enter in nuclear structure probe energy / momentum scales up to order of the vector ( @xmath0)-meson mass . in the spirit of effective field theory , a natural procedure is then to introduce the @xmath0 field explicitly . a powerful and perhaps the most predictive way is to resort to the hidden local symmetric lagrangian which is gauge - equivalent to non - linear @xmath58 model@xcite
. the pion field of the sigma model @xmath60 has the redundancy @xmath61 with @xmath62 .
this redundancy can be exploited to elevate the energy / momentum scale to the vector meson mass scale by gauging the vector @xmath0 .
this leads to hidden local symmetry with the kinetic energy generated dynamically by loop corrections@xcite .
there is magic in this approach if one assumes that the vector meson mass can be taken
it is not understood how the magic comes about , but when treated to the leading order ( at which calculations can be done easily ) it gives some remarkable results such as the ksrf relation ( to all - loop order ) , vector dominance etc . which are difficult to obtain in standard chiral perturbation approaches .
associated with the magic that goes with light " vector meson , there is one specially distinctive prediction which is not present in other approaches , namely , at some high density ( and perhaps also at high temperature ) which may precede possible deconfinement , the @xmath0-meson mass approaches zero as the critical density ( to be denoted @xmath63 ) is approached .
a renormalization - group analysis@xcite showed that as the quark condensate is dialed to zero , the @xmath0 mass which is given by the ksrf formula @xmath64 goes as m_~g_0 .
it should be noted that it is not the pion decay constant but the hidden gauge coupling @xmath65 that controls the property of the @xmath0 mass .
this can be understood with hidden scale symmetry ( discussed below ) : .
the @xmath0 mass is scale - invariant in the vacuum and is expected to remain scale - invariant in medium .
the point at which the mass vanishes is referred to as vector manifestation ( vm ) fixed point . " as stressed in @xcite , the vm fixed point is _ not _ in qcd proper .
it may be not only _ hidden _ but also _ non - reachable _ in qcd in the vacuum .
however it is suggested that it can be generated in highly dense medium .
it will turn out in what follows that the approach near to the vmfp which may be different from the chiral symmetry restoration density @xmath44 can play a crucial role in compact - star matter .
another hidden symmetry highly relevant to nuclear dynamics is global scale symmetry , presumably connected to a low - mass scalar @xmath58 .
qcd at classical level is scale - invariant in qcdlite , " i.e. , @xmath66 for @xmath67 for non - strange nuclear interactions and also @xmath68 if strangeness is involved , but the scale symmetry is broken quantum mechanically by the qcd trace - anomaly .
it is renormalization - group invariant , so can not be turned off in qcd dynamics .
however similarly to the @xmath69 anomaly , it may be restored or rather made emerging " in medium , particularly at high density .
it is this possibility that i will explore for probing the source of the proton mass .
a scalar meson of mass @xmath70 mev has been playing a very important role in nuclear physics , in , say , nuclear forces , relativistic mean - field models etc .
but it has been totally unclear what that scalar is in qcd@xcite . in particle physics ,
the low - lying " scalar higgs scalar @xmath71 poses a similar puzzle . in one of the candidate scenarios for higgs , it is identified as a ng boson associated with broken scale symmetry with an infrared ( ir ) fixed point for @xmath72@xcite .
although difficult to identify , it is surprisingly simple to see where that scalar is hiding " in the strong interactions .
the simplest description was given by yamawaki for higgs physics and going beyond the standard model@xcite .
the reasoning made in the strong interactions is quite similar as the sm higgs lagrangian is known to be equivalent to this linear sigma model .
let us start with the linear sigma model for flavor @xmath73 with the quartet of scalars ,
i.e. , the triplet of pions @xmath34 and a scalar @xmath13 .
( i shall avoid the notation @xmath58 customarily used in the literature for linear sigma model .
it is reserved for the dilaton . )
there is one parameter , @xmath74 , in the model which can be dialed to strong coupling limit ( @xmath75 ) or to weak coupling limit ( @xmath76 ) . in the strong coupling limit
, one obtains the usual nonlinear sigma model with the scalar @xmath13 absent , hence no scale symmetry . the usual current algebra term embodying low - energy strong interactions
is of course not scale - invariant .
standard chiral perturbation theory with baryons explicitly implemented is based on this framework . on the other hand ,
the weak coupling limit @xmath76 gives rise to a scale - invariant nonlinear sigma model with scale symmetry broken only in a potential that encodes the trace anomaly of qcd . in the dilatonic higgs model of @xcite ,
the higgs mass can be small because it is near the ir fixed point .
now the central point of this article : _ in nuclei and nuclear matter , by dialing density , one can sample from the strong coupling regime to the weak coupling regime . _ in the strong coupling limit which is applicable to low - energy hadronic interactions in the matter - free vacuum , nonlinear sigma model works , with the fourth component of the chiral four vector banished to infinity ( say @xmath77 gev ) .
however in nuclei and nuclear matter at @xmath78 , one knows from walecka - type relativistic mean - field ( rmf ) model that a scalar of @xmath79 mev is needed , supplying the attraction that binds nucleons .
what is needed here is not @xmath13 that figures in the gell - mann - oakes - renner ( gmor ) linear sigma model but a chiral singlet with little mixing into the fourth component field . near nuclear matter density
, there is gas - liquid phase transition .
i wo nt deal with it , focusing entirely on @xmath80 . ] , the relevant degrees of freedom are nucleons , a scalar of @xmath79 mev and the vector mesons @xmath0 and @xmath59 of @xmath81 mev treated at the mean field ( with pions entering at loop order ) giving rise to relativistic mean - field theory ( rmft ) which is roughly equivalent to landau fermi liquid theory . here
there is no scalar corresponding to the fourth component of the chiral four - vector . as density
is increased above @xmath82 , however , it is expected that the pion and a scalar will start to approach closer to the structure of linear sigma model , ultimately forming the massless quartet ( in the chiral limit ) .
such a scenario was first proposed by beane and van kolck@xcite , and then given a support with a scale - invariant hidden local symmetry model@xcite .
the limit at which the triplet of pions and scalar form the degenerate massless quartet is referred to as
dilaton limit fixed point ( dlfp ) .
" there is nothing to suggest that the dlfp coincide with the vm fixed point .
one possibility@xcite that has been recently studied is that the dlfp is connected to an infrared fixed point ( irfp ) in qcd that was proposed by crewther and tunstall ( referred to as ct)@xcite and golterman and shamir@xcite .
both postulate the presence of an ir fixed point at which scale ( conformal ) invariance is restored and its spontaneous breaking gives rise to a pseudo - nambu - goldstone boson ( in the presence of the trace anomaly ) .
the characteristic of the ir fixed point is different in the two approaches , but at the leading order of the combined scale symmetry and chiral symmetry , termed scale - chiral symmetry , they are quite similar@xcite . here
i will resort to ct .
there is a great deal of controversy as to whether such an ir fixed point exists in qcd for @xmath83 that concerns us in nuclear physics .
some of this matter are discussed in the references cited above , including @xcite .
the point i will make here is that regardless of whether such an ir fixed point exists in the chiral - scale limit in qcd , scale symmetry is relevant in nuclear physics and connected to the possible source of proton mass , can be probed in nuclear physics .
i now turn to applying the theoretical framework developed above to nuclear phenomena .
i will take the effective lagrangian that i will refer to as scale - invariant hls " that i will consider first without explicit nucleon fields ( @xmath68hls for short ) and then with nucleon fields ( @xmath84hls ) .
when skyrmions are put on crystal lattice , one striking phenomenon observed is a cusp in the symmetry energy @xmath85 defined in the energy per particle @xmath86 of baryonic matter@xcite e(n , x)=e(n,0)+sx^2 + [ s ] where @xmath87 with @xmath88 and @xmath89 is the baryon number density . as shown in fig.[cusp ] , as skyrmions fractionize into half - skyrmions at @xmath90 , the cusp ( fig.[cusp ] ( right panel ) ) is formed in such a way that the @xmath91 first decreases and then goes up .
this cusp is a generic phenomenon involving the pion field only associated with the topology change .
the density at which the cusp is located , however , is determined by the other degrees of freedom than the pion figuring in the lagrangian .
while the precise value for @xmath92 is difficult to determine , with the full @xmath68hls lagrangian , it is reasonably located at @xmath93 .
i will take this value in what follows .
the numerical details are not very sensitive to the exact location as long as it it near @xmath94 . at @xmath92@xcite .
, title="fig:",height=188 ] at @xmath92@xcite .
, title="fig:",height=188 ] what produces this cusp on crystal lattice is the topology change from skyrmions to half - skyrmions at @xmath92 with the vanishing chiral condensate @xmath95 in the half - skyrmion phase .
i will discuss below what this implies in the nuclear tensor force as the density @xmath96 is approached from below .
another surprising observation at @xmath92 is that the effective mass of the nucleon in the skyrmion medium reveals a hidden parity - doublet symmetry with a nonzero mass @xmath97 as @xmath98@xcite m_n = m_0 + m()[mass ] with @xmath99 as @xmath98 at @xmath100 .
since pions are still present in the half - skyrmion phase , @xmath98 may not be interpreted as the restoration of chiral symmetry .
nonetheless @xmath97 can be considered as a component of mass that is chirally invariant .
this indicates that parity doubling in the baryon structure is generated " by the topology change .
in fact this can be matched to a chirally - invariant mass in the parity - doubled nucleon model where @xmath97 is put _
ab initio _ into the lagrangian as was originally done by detar and kunihiro@xcite .
when the @xmath68hls lagrangian that gives rise to the mass formula ( [ mass ] ) on crystal lattice is coupled scale - chiral invariantly to parity - doubled nucleons with a chirally invariant mass @xmath101 in a continuum @xmath84hls , one obtains the same mass formula as ( [ mass ] ) in dense medium@xcite with @xmath101 replacing @xmath97 in ( [ mass ] ) and @xmath102 replacing @xmath103 . in this continuum model , @xmath98 is replaced by @xmath41 , i.e. , chiral symmetry is restored , with the baryons parity - doubled . in both the skyrmion crystal and the baryon parity - doubled model , @xmath104 which stays constant for @xmath105 for the former and for @xmath106 for the latter where @xmath107 is the density at which the parity - doubling takes place .
it is likely that @xmath108 .
what is significant for my purpose is that @xmath109 could be large , say , @xmath110 .
this is not directly associated with the nambu mechanism anchored on the spontaneous breaking of chiral symmetry .
the mass @xmath109 is a chiral scalar .
this is apparently unrelated to the @xmath40 generated by the nambu mechanism .
so where does this chiral scalar mass come from ? how is this related to the mass generated by the cheshire cat mechanism i started with ?
this is the question nuclear physicists would like to find the answer to .
as noted above , the skyrmion model and then quasiquark model for the nucleon are equivalent in the large @xmath36 limit .
the quasiquark model unifies baryons and mesons and works very well when corrected with certain @xmath111 corrections for various static and dynamic properties .
the origin of mass for both the proton and the @xmath0 is the constituent quark mass given by the dynamically generated quark condensate .
especially , their masses have a simple ratio , e.g. , @xmath112 . now
the skyrmion approach is also a unified approach as skyrme suggested in his seminal paper and more generally in @xmath84hls in that it involves one unified " lagrangian that accounts for both mesons and baryons .
but there is no equivalent to constituent quark in the skyrmion description .
attempts to obtain a constituent @xmath111 skyrmion , called
qualiton"@xcite , have thus far been unsuccessful@xcite .
there are therefore no such simple mass relations between the mesons and baryons as in the quasiquark model .
in fact i see no logical connection between the two in the cheshire cat scenario as will be specified below .
furthermore a tension arises between the skyrmion picture and the quasiquark picture when we go to dense matter . while the proton mass moves to a constant @xmath113 after shedding off @xmath114 subject to the decrease in the quark condensate , the @xmath0 mass which
is given in medium by @xmath115 goes to zero because @xmath116 goes to zero at the vm fixed point .
@xmath116 is not directly related to the quark condensate .
this dichotomy would imply the scenario for the mass as given here is drastically different from the * q * mechanism in light quark hadrons .
if one were to endow a chiral - invariant mass @xmath117 to a @xmath16 in the proton as one would do for the parity - doubling in the @xmath16 model , it would preserve the simple ratio @xmath118 and other constituent - quark properties in medium as it would in the vacuum
. however it would be at odds with the vm for the vector meson .
furthermore in this picture , there would be a tremendous increase in excitations of both parities in the spectra of both baryons and mesons in medium , even at normal matter density which is accessible in the laboratories .
this possibility seems to be ruled out as there is no evidence for such an enhancement .
0.3 cm @xmath43 * t(opology ) scenario * : in approaching the vector manifestation ( vm ) fixed point in @xmath84hls implemented with topology change , @xmath119 ( vm and scale - invariance ) but @xmath120 ( by scale symmetry ) , hence @xmath121 .
i suggest that this scenario is favored in nuclear physics .
let me now turn to what nature says about the issues discussed above and what one can infer from them .
the cusp observed above in the symmetry energy given by the skyrmion crystal can be reproduced by the behavior of the tensor forces in nuclear interactions .
it is due to the fact that nuclear symmetry energy is dominated by the tensor forces . in terms of the lagrangian
used in this paper , @xmath84hls , the ( total ) tensor force is given by one - pion and one-@xmath0 exchanges . in calculating the effective forces ,
one is to take into account the coupling constants and masses of the mesons involved in the exchanges that are scaling with density .
the density dependence is prescribed by both the intrinsic density dependence ( idd for short ) inherited from qcd at the scale at which the correlators of qcd and effective theory are matched and the in - medium renormalization group properties of the effective field theory , @xmath84hls .
what s involved here are two things .
first the @xmath34-exchange tensor force and the @xmath0-exchange tensor force come with opposite signs so they tend to cancel in the range of forces entering in the pertinent nuclear interactions .
the pion mass and coupling constants are protected by nearly exact chiral symmetry , consequently the pion tensor is more or less unaffected by density . on the other hand ,
the @xmath0 tensor increases in strength with the @xmath0 mass falling as density increases toward @xmath92 and then gets quenched strongly by the hidden gauge coupling @xmath116 going toward zero as the vm fixed point is approached .
the net effect is then that the sum of the two tensor forces decreases due to the cancelation between them up to @xmath92 and then is taken over by the pion tensor with the @xmath0 tensor suppressed , hence increasing beyond @xmath92 .
this property is seen in the left panel of fig.[tensor ] . and
@xmath0 tensor forces vs. @xmath122 ( fm ) for different denstity @xmath89 .
the @xmath0 tensor is completely suppressed at @xmath123 .
right panel : the symmetry energy @xmath91 predicted by the @xmath124 with @xmath84hls with @xmath125 .
the eye - ball slope change is indicated by the colored straight lines .. ,title="fig:",height=207 ] and @xmath0 tensor forces vs. @xmath122 ( fm ) for different denstity @xmath89 .
the @xmath0 tensor is completely suppressed at @xmath123 .
right panel : the symmetry energy @xmath91 predicted by the @xmath124 with @xmath84hls with @xmath125 .
the eye - ball slope change is indicated by the colored straight lines ..
,title="fig:",height=222 ] now the net effect of the tensor forces on the symmetry energy in nuclear matter near @xmath82 can be approximately expressed in the form that comes from the closure approximation @xmath126 where @xmath4 is a dimensionless constant .
this simple form follows from the fact that the tensor force predominantly excites from the ground state to excited states peaked at @xmath127 mev .
this expression leads to the @xmath91 that decreases as @xmath89 approaches @xmath92 and then increases as @xmath89 goes above @xmath92 .
this is precisely the cusp structure predicted by the skyrmion model .
it should be mentioned that this behavior of @xmath91 supports the scaling behavior of the @xmath84hls lagrangian described above in the tensor force channel .
it is important to note that what plays a crucial role here is the vm fixed point toward which dense matter flows .
it is fairly well recognized that the symmetry energy @xmath91 is the most important ingredient in the equation of state ( eos ) for massive compact stars . and
it is also the most poorly known at high density .
thus the tensor force is the key ingredient for compact - star physics yet to be figured out .
if the scenario described above is correct , a reliable approach to the eos involves an intricate combination of topology change and in - medium properties of nucleons and mesons , specially the @xmath0 meson .
it should of course be recognized that the symmetry energy @xmath91 in nuclear matter is more involved than just what the tensor force gives .
it can get contributions from different components of the nuclear force .
in fact at low density , the closure - approximated expression for @xmath91 can not be reliable .
this is because first of all , the topology structure inferred from skyrmion crystal can not be trusted at low density . furthermore it is known that the cluster structure of matter can be important at low density .
nonetheless the behavior for @xmath128 is expected to be trustful as it is mainly controlled by the vm fixed point .
a full rg flow calculation using @xmath124 potential described below gives the symmetry energy given in fig.[tensor ] ( right pannel ) .
the cusp structure , smeared by other components of the nuclear force , gives rise to a distinctive changeover from a soft @xmath91 to a stiffer @xmath91 at @xmath92 .
this stiffening accounts for @xmath129 solar mass compact star recently observed .
exploring the behavior of the tensor force in nuclei is a fertile and challenging field in nuclear physics .
i give a few illustrations as to where it stands .
the long lifetime for the gamow - teller transition in @xmath130c presents a case where the structure of the tensor force is exposed at a matter density near @xmath82@xcite .
it sheds light on complex and highly intricate interplay between the pion tensor force and the @xmath0 tensor force at matter densities near @xmath82 . as shown in fig.[tensor ] ( left panel ) and discussed above , up to @xmath131 , the net tensor force at the range relevant to the process decreases as density increases , nearly vanishing at @xmath96 .
this cancelation manifests spectacularly in the gamow - teller matrix element involved in the transition @xmath130c @xmath132 @xmath130n because the gt matrix element nearly vanishes as density approaches @xmath82 around which the wave functions overlap .
what is significant in this process is that the transition involved samples the density regime where such cancelation takes place and the energy spectrum in @xmath130n for the corresponding states comes out in perfect agreement with experiments .
this provides a consistency check between the weak current and the hamiltonian , both of which scale with @xmath133 which , in scale - chiral effective theory , is locked to @xmath134 .
it should be pointed out clearly that the cancelation in the transition matrix element , although consistent in a nontrivial way with the hamiltonian giving the energy spectrum , involves an interplay of complicated many - body correlations and can not single out unequivocally the role of the tensor force subject to the scaling parameters .
one must admit that the c-14 dating is at best a tenuous evidence for the effect one is looking for .
one would like to zero - in on processes that directly exhibit the scaling behavior of the tensor force as a function of density .
this will require precision in both theory and experiments .
it would be mandatory to first search for experiments that can be done and analyzed with precision .
this is a challenge for theorists and experimentalists . from the theory side ,
there is one promising development , both surprising and intriguing .
it is that the tensor force effective in nuclear interactions appears to be a fixed - point interaction in the sense of landau fermi - liquid fixed point theory . to be practical and concise it is clearest to put it in terms of the @xmath124 potential@xcite in the renormalization - group approach to effective theory for nuclear systems . in the wilsonian rg approach
, one can adopt what s called double - decimation rg"@xcite .
@xmath124 is the effective potential arrived at from the first decimation with a cutoff at @xmath135 @xmath136 .
it is applicable for the vacuum as well as in medium , endowed with suitable density dependence as described above .
the second decimation brings it to the landau fermi - liquid fixed point and correlations on top of the fermi surface .
it is not proven but observed in numerical results that the tensor force appears to be a fixed - point interaction . to clarify what s involved , let s think of approaching nuclear many - body problems using @xmath124 potentials constructed from @xmath84hls lagrangian .
this is a natural approach anchored on nuclear effective field theory exploiting rg decimations . focusing on the tensor force , i denote by @xmath137 the tensor potential given by the exchanges of @xmath34 and @xmath0 implemented with idd defined at a scale relevant to nuclear calculations , roughly about the vector meson mass . in the framework
adopted , this potential is applicable both in the vacuum and in medium . in the vacuum
it will be @xmath138 .
denote by @xmath139 the low - momentum component potential obtained from the first decimation defined with the cutoff @xmath135 @xmath136 which seems to work well , corresponding approximately to the momentum scale up to which precision data are available for scattering data .
the first observation is made in the matter - free space , @xmath140 .
for this , no separation into double decimations is needed .
consider the two - nucleon matrix element in the tensor channel @xmath141-@xmath142 with @xmath143 and @xmath144 .
the surprising result shown in fig.[tensor - free ] is that f |v^t_lowk|i f |v^t|i and the beta function ( v^t_lowk)0 . and @xmath145 in matter - free space .
here the potential is the bonns potential , approximating the leading - order contribution from the @xmath84hls lagrangian.,title="fig:",width=264 ] and @xmath145 in matter - free space . here
the potential is the bonns potential , approximating the leading - order contribution from the @xmath84hls lagrangian.,title="fig:",width=264 ] this result suggests that the tensor force given at the matching scale is unrenormalized by strong interactions in the rg flow in matter - free space .
why the tensor force is at its fixed point , with the @xmath146 function nearly vanishing , has at present no explanation .
let me go even further and look at processes in nuclei with density @xmath147 , which requires the double decimation including higher - order correlations .
the c-14 dating case offers no clear hint due to the intricate interplay between the tensor force with other components of the nuclear force .
fortunately there is a case , known up to date , relevant to the issue which hints at the direction to take .
it is shell evolution in exotic nuclei .
it has been suggested that one of the underlying causes for shell evolution is based on the strong monopole part of effective nucleon - nucleon interactions for which the tensor force is found to play an important role in the appearance or disappearance of the spherical magic numbers in light - mass nuclei , and for heavier nuclei , in the observed drift of single- particle levels along the @xmath148 isotonic and the sb ( @xmath149 ) isotopic chains@xcite .
the monopole " matrix element of the two - body interaction between two single - particle states labeled @xmath150 and @xmath151 and total two - particle isospin @xmath152 is given by @xmath153 .
remarkably this matrix element affects the evolution of single - particle energy ; @xmath154 where @xmath155 represents the change of the single - particle energy of protons in the state @xmath150 when @xmath156 neutrons occupy the state @xmath151 .
it turns out that the matrix elements @xmath157 and @xmath158 have opposite sign for the tensor forces if @xmath150 and @xmath151 are spin - orbit partners .
let me summarize the salient features of the shell evolution connected to the tensor force found by otsuka@xcite .
the phenomenological potential av18 and the one given by chpt at n@xmath159lo , both treated la @xmath160 and also other realistic potentials ( such as bonns ) are found to give the same results .
the cutoff @xmath161 is varied and the result is found to be independent of the cutoff around 2.1 @xmath162 . shown in fig.[pfshell ] is the shell evolution in the pf and sd regions calculated by including high - order correlations using the q - box formalism to 3rd order@xcite . while the central part of the potential is strongly renormalized by high - order terms , the tensor forces are left unrenormalized , leaving the bare " tensors more or less intact .
interaction for ( a ) t=0 and ( b ) t=1 . from @xcite.,width=302
] what the result shows is that the sum of the short - range correlation and medium effects as taken into account by the 3rd order q - box@xcite leaves the bare tensor force unchanged .
given the robustness of the result , one may take this result to imply that up to nuclear matter density v_low k^tensor=([v_low k^tensor ] , ) 0.[beta ] this is a surprisingly intriguing result .
if one were to take this result seriously , it would mean that the tensor force is unrenormalized both in the free space and in medium .
the q - box used here corresponds to high - order correlation effects in nuclear medium , thus we arrive at the observation that highly involved nuclear correlations also cancel in the tensor force .
however the analysis made in @xcite contains no idd , hence does not contain the information on the decrease of the tensor force in density up to @xmath78 which plays an important role in the c14 dating .
there are no experimental data yet to check whether the results in @xcite confirm or rule out the density dependence in the tensor force predicted . the fixed - point property ,
if valid , should apply with the idd incorporated .
it is a challenging problem to confront the idd - implemented tensor force in the monopole matrix elements figuring in the shell evolution .
this would require both highly sophisticated precision calculations and precision experimental data .
the scaling of the mass ratio @xmath163 is expected to set in as the density surpasses @xmath93 and becomes more prominent as it goes toward the vm fixed point .
this will be an indication for difference in the sources for the masses of baryons and mesons .
the rib machines , both in operation and also in project , could reach density @xmath164 and hence probe the regime where the changeover of the state of matter takes place .
what can be studied in rib and going beyond to what could be offered by more powerful machines like fair of darmstadt will confront what will be observed in gravity waves coming from the merging of neutron stars .
the gravity waves are estimated to provide information of density of order @xmath165 at temperatures of order @xmath166 mev@xcite .
the onset of the mass ratio @xmath167 triggers the change from soft to hard in the symmetry energy @xmath91 in compact stars as illustrated in fig.[tensor ] .
it will be extremely interesting to observe such changeover in the tidal deformability in compact stars as observable in gw .
this possibility is a hot topic in nuclear astrophysics@xcite .
i discussed two possible scenarios for the origin of masses of light - quark hadrons , specifically the proton and the @xmath0 meson . as the baryonic matter density
is increased beyond @xmath168 but before the particle picture breaks down , by e.g. , possible percolation transition , the mass ratio @xmath163 could behave in different ways , signaling how the hadrons shed their masses : * * q mechanism " * : @xmath169 .
going toward the chiral restoration density @xmath44 , @xmath46 and @xmath45 .
this follows from the nambu mechanism applied to quasiquarks with the assumption that the quasiparticle description continues to hold .
here hidden symmetries play no manifest role . *
* t mechanism " * : @xmath170 .
going toward the vm fixed point @xmath63 which may or may not coincide with @xmath44 , @xmath46 and @xmath171 . the former comes about because the @xmath0 mass is nearly scale - invariant@xcite and governed by the vm fixed point and the latter is controlled by ( broken ) scale symmetry . in both , the cheshire cat mechanism with topology change , hidden symmetries
with parity doubling and the vector manifestation and the special structure of the nuclear tensor forces play an essential role . the emergence of scale symmetry , hidden in nature , could take place _
after _ the vm fixed point and approaching the dlfp .
deconfinement does not figure , at least up to compact - star - matter densities .
on the theory side , there are quite a few questions to be answered . among them
, regarding the half - skyrmion on crystal lattice , could it be related to the chiral bag at the magic angle at which the baryon charge fractionizes into two halves ?
if so , how is it related to the topology change that plays an intriguing role for the tensor force and ultimately in the eos for dense matter ?
another issue is a possible cheshire cat interpretation of the hadron - quarkyonic transition@xcite which resembles the skyrmion - half - skyrmion transition discussed in this note and detailed in@xcite .
finally the topology change could provide the additional attraction needed in the akaishi - yamazaki mechanism for high density kaonic - proton matter@xcite . in @xcite
, the additional attraction was attributed to what was called vacuum clearing " in ref @xcite . in @xcite , it was found that the changeover from skyrmions to half - skyrmions on crystal lattice gave rise to a substantial attraction over and above the wess - zumino term which is equivalent to the weinberg - tomozawa term used in @xcite coming from the dilaton associated with the emergent scale symmetry .
the scale - invariant hidden local symmetry scheme developed above could give a reliable value for this attraction .
y. l. ma , m. harada , h. k. lee , y. oh , b. y. park and m. rho , `` dense baryonic matter in conformally - compensated hidden local symmetry : vector manifestation and chiral symmetry restoration , '' phys .
d * 90 * , no . 3 , 034015 ( 2014 ) ; m. harada , y. l. ma , h. k. lee and m. rho , `` fractionized skyrmions in dense compact - star matter , '' in@xcite ,
arxiv:1601.00058 [ nucl - th ] .
d. b. kaplan , `` qualitons , '' nucl . phys .
b * 351 * , 137 ( 1991 ) ; j. r. ellis , y. frishman and m. karliner , `` constituent quarks as solitons , '' phys .
b * 272 * , 333 ( 1991 ) ; p. h. damgaard and r. sollacher , `` qualitons from qcd , '' phys .
b * 322 * , 131 ( 1994 ) .
j. w. holt , g. e. brown , t. t. s. kuo , j. d. holt and r. machleidt , `` shell model description of the c-14 dating beta decay with brown - rho - scaled nn interactions , '' phys .
lett . * 100 * , 062501 ( 2008 ) . for an up - to - date review and
references on @xmath124 and @xmath16-box folded diagram technique mentioned in the text , see t. t. s. kuo , j. w. holt and e. osnes , `` introduction to low - momentum effective interactions with brown - rho scaling and three - nucleon forces , '' phys .
scripta * 91 * , no . 3 , 033009 ( 2016 ) .
t. otsuka , t. suzuki , m. honma , y. utsuno , n. tsunoda , k. tsukiyama and m. hjorth - jensen , `` novel features of nuclear forces and shell evolution in exotic nuclei , '' phys .
lett . * 104 * , 012501 ( 2010 ) .
n. tsunoda , t. otsuka , k. tsukiyama and m. hjorth - jensen , `` renormalization persistency of tensor force in nuclei , '' phys .
c * 84 * , 044322 ( 2011 ) .
l. mclerran and r. d. pisarski , `` phases of cold , dense quarks at large @xmath36 , '' nucl .
a * 796 * , 83 ( 2007 ) ; k. fukushima and t. kojo , `` the quarkyonic star , '' astrophys . j. * 817 * , no . 2 , 180 ( 2016 ) . | i describe a possible scenario for the origin of proton mass in terms of cheshire cat , half - skyrmions , topology change and interplay between hidden chiral - scale symmetry and induced local symmetry .
this differs from the standard constituent - quark scenario .
as the baryonic matter density is increased toward the vector manifestation ( vm ) fixed - point at which the @xmath0 mass is to vanish , the effective in - medium mass ratio @xmath1 is to tend to zero proportionally to @xmath2 where @xmath2 is the in - medium hidden gauge coupling constant .
i develop the thesis that the intricacy involved in the mass generation could be decoded from experiments at rib accelerators and massive compact stars . |
uncertainty relations can be established in any physical theory involving random conjugated variables characterized by probability distribution functions . the present paper deals with uncertainty relations in statistical mechanics .
statistical mechanics combines equilibrium thermodynamics with a probability postulate @xcite . following the standard gibbs ensemble theory , in the microcanonical ensemble it is assumed that all mechanical variables are constrained to a fixed value ( or are allowed to vary within a fixed small range ) , while their entropy - conjugated variables are free to fluctuate .
other ensembles ( canonical , grand - canonical , ... ) are introduced by relaxing the constraint on one or more mechanical variables .
this is done by assuming that these mechanical variables can be exchanged with a reservoir , which is defined as a macroscopic body of dimensions much larger than those of the system .
the exchange is supposed to be controlled by parameters ( which are characteristic of the reservoir ) that are identified as intensive variables .
formally , the change of variables when passing from the microcanonical to other ensembles is undertaken via legendre transforms . in the spirit of equilibrium thermodynamics ,
the intensive variables are assumed to have the same value both in the reservoir and in the system and are usually treated as non - fluctuating variables .
therefore , within this basic framework , uncertainty relations are not expected to apply @xcite .
while it seems reasonable that intensive variables of a very large reservoir ( a system characterized by large amounts of mechanical variables ) are treated as non - fluctuating variables , there is no reason for such an assumption when dealing with intensive variables corresponding to a small reservoir or system .
therefore , we will adopt the point of view that , in general , for a system in contact with a reservoir ( regardless of size ) fluctuations of intensive variables should occur in the system induced by the fluctuations of the corresponding conjugated mechanical variables . from this point of view , uncertainty relations are expected to play an important role for a correct understanding of the behaviour of equilibrium fluctuations of thermodynamic variables , with consequences in the experimental measurements of these variables .
actually this simple picture has received some experimental support @xcite and thus it is reasonable to ask whether or not uncertainty relations can be formulated and what their applicability limits are .
the problem was already discussed by bohr and heisenberg as soon as heisenberg proposed the _ uncertainty principle _ in quantum mechanics @xcite .
therefore , during the last century a number of papers were published dealing with this problem .
an interesting article where these issues have been reviewed is the paper by uffink and van lith @xcite .
more recently , velzquez and curilef @xcite revisited the problem within the framework discussed some years before by mandelbrot @xcite , gilmore @xcite and others .
the vast majority of these works have been formulated from a purely theoretical point of view , and applications to particular examples ( analytical or numerical ) are scarce . the aim of the present paper is to numerically analyze the validity of uncertainty relations in two simple lattice models and shed some light on controversial interpretations @xcite the paper is organized as follows . in section [ sec : relations ] we summarize a derivation of the uncertainty relations in statistical mechanics . in section [ sec : examples ] we numerically analyze two examples corresponding to an ising paramagnet and an ising ferromagnet .
finally , in section [ sec : conclusions ] we discuss the relevance of the results in relation to temperature measurements in small thermodynamic systems and draw some conclusions .
following the mathematical treatment proposed in ref .
, let us consider a continuous probability distribution , @xmath0 , defined within the domain @xmath1 , which is assumed to vanish at the boundary @xmath2 .
the average value of the derivative of any analytical function @xmath3 with respect to one of the components of @xmath4 , can be expressed as : @xmath5 where we have defined the affinities : @xmath6 two interesting particular cases of the result in ( [ eq : result1 ] ) can be obtained by choosing @xmath7 and @xmath8 . in the first case one gets @xmath9 and in the second case @xmath10 let us now define the differences : @xmath11 @xmath12 the variances @xmath13 and the covariances @xmath14 of arbitrary differences @xmath15 and @xmath16 can be interpreted as a scalar products in a vectorial space where the metric is given by the probability distribution . in particular ,
the square root of the variance @xmath17 of a random variable @xmath18 ( which we will denote as the fluctuation of @xmath18 ) is the norm of the vector @xmath15 in this metric .
therefore , in our case , taking into account the schwartz inequality , we can write : @xmath19 from ( [ eq : term1 ] ) and ( [ eq : term2 ] ) , the right - hand - side term can be expressed as : @xmath20 thus , we obtain an inequality relating the variances of the variable @xmath21 and the affinity @xmath22 : @xmath23 it is worth noting that other authors have used different mathematical approaches @xcite to treat this problem . in this subsection
we will discuss how to apply the results obtained above in order to establish uncertainty relations in statistical mechanics .
actually , the key point will be to find a convenient definition of the set @xmath24 , the probabilities @xmath25 and the affinities @xmath26 .
we proceed by dividing an isolated universe into two interacting parts that can exchange one or more mechanical variables .
one part ( usually the smallest one , which we wish to study ) will be called the `` system '' , while the other part will be the `` reservoir '' . for the closed universe we assume boltzmann equiprobability and therefore its entropy will be given by : @xmath27 where @xmath28 is the total number of microstates of the universe and @xmath29 is the boltzmann constant .
let us consider the set @xmath24 of mechanical variables describing the system @xmath30 .
the corresponding variables for the universe @xmath31 are constant ( closed ) .
therefore , for the reservoir we have @xmath32 .
from knowledge of the number of microstates of the universe @xmath33 for which the system variables take values within @xmath34 we can determine the probability density : @xmath35 where @xmath36 is an elementary volume in @xmath24 that will be irrelevant for further developments .
it is introduced here in order to have consistent units . assuming that this probability density @xmath37 is a finite function , eq .
( [ eq : distribution ] ) can be written in terms of a _ universe dimensionless entropy _ which depends on the system variables : @xmath38\ ] ] therefore we have @xmath39 which is none other than einstein s formula for fluctuations @xcite . at this point , without making any assumptions about the separability of the function @xmath40 , we can define the affinities between the system and reservoir as the partial derivative : @xmath41 note that @xmath26 has units inverse of those of @xmath42 since @xmath43 is dimensionless . in some cases
, it can be assumed that @xmath40 can be split into a part related to the entropy of the studied system , @xmath44 , and a part related to the reservoir , @xmath45 , so that @xmath46 since functions @xmath44 and @xmath45 only depend on the internal structure of the corresponding parts , it then seems natural to define intensive variables for the system and reservoir in analogy with thermodynamics : @xmath47 therefore , assuming separability , the affinity can be expressed as : @xmath48 the results derived in subsection [ subsec : math ] can now be written in terms of the preceding definitions . without assuming separability , from eq .
( [ eq : term1 ] ) we get : @xmath49 and from the inequality relation in ( [ eq : resultatbo ] ) , we can write : @xmath50 when separability holds , equations ( [ eq : equilibri ] ) and ( [ eq : resultatest])become @xmath51 and @xmath52 respectively .
these are the main results that we will numerically test in the examples in section [ sec : examples ] .
note that for the particular case of choosing energy as a mechanical variable , in the separable case , the above two equations read : @xmath53 and @xmath54 where we have used the more common symbols @xmath55 and @xmath56 to represent the intensive variables conjugated for the energy of the system and reservoir .
equations ( [ eq : equilibri ] ) , ( [ eq : equilibri2 ] ) or ( [ eq : betes ] ) are essentially probabilistic versions of the zero - th principle of thermodynamics @xcite . because of the fluctuations of the system and reservoir , the equilibrium conditions are not an absolute equality , but only apply to statistical averages of the intensive variables .
it is important to note that with the separability assumption the conjugated variables of a system appear as functions of extensive variables of the same system ( as in thermodynamics ) and that the same happens when the roles of the system and reservoir are interchanged .
equations ( [ eq : resultatest ] ) , ( [ eq : resultatest2 ] ) or ( [ eq : energ ] ) state that fluctuations in conjugated thermodynamic variables are related in a form which is similar to the uncertainty relations in quantum mechanics .
these results are usually referred to as uncertainty relations in statistical physics and , even though their interpretation in quantum mechanics is different , they both come from the same mathematical base , as noted in ref . .
at this point it is convenient to point out that the previous results were obtained after neglecting the contribution from the boundary @xmath2 in eq .
( [ eq : result1 ] ) and after considering a continuous character for the variables @xmath42 describing the system .
therefore , the uncertainty relation ( [ eq : resultatest ] ) and the equilibrium condition ( [ eq : equilibri ] ) may not hold if states with extreme values of the variables @xmath42 ( at the boundary @xmath57 ) are likely to occur .
such states will be called boundary states and in the following examples , that have a discrete character , it will be important to evaluate their probability and check under which circumstances they become too high .
let us consider a universe in which the dimensionless entropy @xmath43 can be separated between the system and reservoir and we focus only on the energy of the system @xmath58 as a mechanical variable .
let us assume that @xmath43 or equivalently , the continuous probability density @xmath59 , is centred about a maximum at @xmath60 .
we also require that the function @xmath43 has no other maxima , which is a typical situation in statistical mechanics , far from phase transition points .
exactly at the maximum we have @xmath61 ( otherwise it would not be extremal )
. fluctuations of @xmath62 about zero are related to how much the system fluctuates near the point @xmath60 .
[ fig1 ] shows a schematic representation of this situation .
the plot above shows the probability of the energy fluctuations .
the entropies of the system , the reservoir and the universe are shown in the middle diagram . at the bottom
we show the derivatives of the entropies corresponding to @xmath55 , @xmath56 and @xmath63 .
the arrows indicate the width of the fluctuations of the energy and the inverse temperatures .
if the probability density becomes narrower , energy fluctuations will decrease , but @xmath43 will have a more pronounced maximum .
this means that the slope of @xmath64 ( blue line in the bottom diagram ) increases and consequently the affinity will have larger fluctuations .
if energy fluctuations are small enough we can expand the function @xmath65 about the point @xmath60 .
a first - order approximation gives : @xmath66 where we have taken into account the fact that @xmath67 is fixed and we have defined : @xmath68 which plays the role of an inverse joint heat capacity .
indeed the plots shown in fig .
[ fig1 ] correspond to such a first - order approximation , where the probability is gaussian , the entropy densities are parabolas and the inverse temperatures and affinity behave linearly . note that within this first - order approximation , condition ( [ eq : betes ] ) yields @xmath69 .
besides , the fact that @xmath70 makes ( [ eq : energ ] ) an equalty ( it is indeed a property of the schwartz inequality ) .
thus , replacing ( [ eq : linealitz ] ) into ( [ eq : energ ] ) , we find @xmath71 hence @xmath72^{-1}\ ] ] if we define the system heat capacity @xmath73 and the reservoir heat capacity @xmath74 as @xmath75 @xmath76 where @xmath77 stands for the equilibrium temperature , we get : @xmath78 which is essentially the same result obtained by lindhart , as discussed in ref . , but seen from within a different framework . at the same order of approximation we can compute : @xmath79 @xmath80 @xmath81 @xmath82 from the previous equations , for an infinite reservoir ( @xmath83 ) we obtain the expected results @xmath84 @xmath85 @xmath86 @xmath87 @xmath88 even though many macroscopic systems can be studied as continuous , their nature is usually discrete , and , therefore , equations ( [ eq : equilibri ] ) and ( [ eq : resultatest ] ) should be seen as approximations .
we will now discuss how to deal with discrete systems .
the main difficulty in the application of equation ( [ eq : resultatest ] ) to discrete systems , is how derivatives should be taken in the definition of intensive variables .
we must consider , not only the problem of finding whether defining forward , backward or central derivatives can affect the result , but we must also take into account the fact that @xmath89 and @xmath90 have different corresponding discrete expressions .
let us discuss two possible definitions .
if we consider the second part of eq .
( [ eq : theaffinity ] ) a possible ( centred ) definition of @xmath62 is @xmath91 which is separable if the probability @xmath92 factorizes .
nevertheless this prescription also raises some problems : since the logarithm goes to minus infinity as probability goes to zero , one must be really careful at the edges of the domain and change the centred prescription to backwards and forward prescriptions .
an alternative definition of @xmath93 is the one suggested from the first part of eq .
( [ eq : theaffinity ] ) : @xmath94 this definition is not separable as a product of terms for the system and the reservoir , not even when the probability factorizes . as a consequence
, the concept of intensive variables can not be thought of as a property of a single system . instead , in this case , we shall deal with affinities that are a property of each pair of systems in equilibrium . in the first example in the next section [ sec : examples ] we will compare the results obtained with the two prescriptions in order to evaluate
how much the choice can affect the results for small systems .
in order to discuss the applicability of the above expressions , we will test them in two numerically treatable examples : a non - interacting lattice system ( paramagnetic ising system ) and an interacting lattice system ( ferromagnetic ising system ) .
we will focus on the validity of expressions ( [ eq : equilibri ] ) and ( [ eq : resultatest ] ) for various universe sizes , various system sizes within each universe and different values for the extensive fixed variables of the universe .
our aim is to deepen our understanding of the validity of these equations .
let us firstly analyze the validity of the formalism developed above for a model without interaction .
we shall consider a universe consisting of @xmath95 non - interacting spins @xmath96 that can take two values @xmath97 , depending on whether they are parallel or antiparallel to an external field @xmath98 ( which is a fixed parameter of the model ) .
the hamiltonian is given by @xmath99 .
since the universe is isolated ( constant energy ) the number of spins @xmath100 in the @xmath101 state and the number of spins @xmath102 in the @xmath103 state ( fulfilling @xmath104 ) are fixed .
therefore , the number of possible configurations of the universe is given by : @xmath105 we now consider a partition of the whole universe into the two sets : the system and the reservoir so that @xmath106 .
we will choose the number of spins @xmath107 in the @xmath101 state in the system as the internal variable of interest in this example ( this choice gives simpler expressions than choosing the energy ) .
its probability function will be given by : @xmath108 where @xmath95 , @xmath100 and @xmath109 are parameters .
note that the variable @xmath107 ranges between the extremes corresponding to its minimum value given by @xmath110 and its maximum value given by @xmath111 . using the probability function ( [ probparamagnet ] ) we can explicitly compute the average value @xmath112 and its variance as @xmath113 for illustrative purposes , in this first example we compute the probability for the system to reach boundary states @xmath114 .
[ fig2 ] shows the value of this probability for two universes of sizes @xmath115 and @xmath116 as a function of @xmath100 and for different values of the system size @xmath109 .
this result shows that the probability of the boundary states departs from zero when the system or reservoir are too small and when the value of @xmath117 is too close to its extremes and this fact constrains the number of available states in the system .
so far there has been no need to take one of the prescriptions described in section [ subsec : discrete ] .
however , in order to perform the study of the conjugated intensive variables we must now choose one of the prescriptions .
the first proposed option , according to eq .
( [ eq : afinitatlog ] ) , enables us to benefit from the separation of the entropy density .
as the spins are non - interacting , when we define the entropy using the logarithm and neglect irrelevant constants , it can be trivially separated as : @xmath118 with @xmath119 and @xmath120 to estimate the intensive variables of the system ( @xmath121 ) and reservoir ( @xmath122 ) we approximate the derivatives of @xmath44 and @xmath123 by a finite difference scheme .
since the variable @xmath107 is bounded , we will use central differences everywhere except at the extremes , where we will use forward or backwards differences .
thus : @xmath124 and @xmath125 fig .
[ fig3 ] shows the difference @xmath126 as a function of @xmath100 , for universes of sizes @xmath127 ( left panel ) and @xmath128 ( right panel ) and different system sizes @xmath109 ( ranging from @xmath129 to @xmath130 ) .
notice that the roles of the system and reservoir are , in this case , symmetric .
as can be seen , the difference @xmath131 vanishes when @xmath132 ( which determines the available energy ) is not close to its upper or lower extremes .
it is clearly observed that the range of validity of the result in eq .
( [ eq : equilibri2 ] ) increases when the system size @xmath109 approaches half the universe size @xmath130 , since we have more available states and therefore the importance of the boundary states on the computed averages decreases .
[ fig4 ] shows the test of the uncertainty relation ( [ eq : resultatest2 ] ) as a function of the total number of up spins @xmath100 . for @xmath115 and @xmath116 , and different values of @xmath109 .
similar conclusions to those observed in fig .
[ fig3 ] can be deduced : when @xmath109 and @xmath100 are far from their boundary values , the uncertainty relation is fulfilled .
only when the fluctuating variable @xmath107 is strongly constrained to a tiny number of states by the system size @xmath109 or the available fixed universe energy @xmath100 , do the surface terms neglected in the derivation of the uncertainty relation play a role and the bound is broken .
alternatively , we can study the second method proposed for discrete systems and define the joint affinity @xmath62 as in eq .
( [ eq : afinitatdisc ] ) . in this case , however , it is not possible to separate @xmath62 into two terms , one corresponding to the system and another one corresponding to the reservoir . numerical computation for the same cases studied before allows us to plot the results shown in figs .
[ fig5 ] and [ fig6 ] . the same qualitative results concerning the validity of ( [ eq : equilibri ] ) and ( [ eq : resultatest ] ) can be concluded .
note that the values of @xmath133 in fig .
[ fig5 ] are smaller than those in fig .
therefore , for discrete systems , the choice of the approximation method discussed in section [ subsec : discrete ] may affect the results from a quantitative point of view , even in such a simple case . in this second example we study the case of a discrete system with interactions . since interaction can induce correlations ,
in this case we should not assume that the entropy density @xmath43 is separable and thus we should restrict ourselves to only testing the validity for equations ( [ eq : equilibri ] ) and ( [ eq : resultatest ] ) .
let us consider a universe consisting of an ising model on a two- dimensional square lattice with size @xmath134 , with nearest - neighbour interactions and periodic boundary conditions .
the system of interest will be a rectangular subset of spins with size @xmath135 . even though a general treatment requires the simultaneous analysis of energy and magnetization fluctuations , in the absence of an external field
, it is enough to analyse only the energy .
the hamiltonian of the universe reads @xmath136 where @xmath137 are the spin variables and the sum extends over all the @xmath138 nearest - neighbour pairs of the universe .
note that the energy of the universe will be discretized in steps @xmath139 and ranges from @xmath140 to @xmath138 for any given configuration , the evaluation of the energy of the system will be done by considering that the energy associated with the bonds along the boundary of the system is equally divided between the system and reservoir .
therefore , the energy of the system is discretized in steps @xmath141 . in order to perform a microcanonical analysis of the averages at constant @xmath67
we have used the following procedure .
we have randomly generated spin configurations and classified them according to the universe energy .
inside the set corresponding to the same value of @xmath67 we have computed the fraction of configurations that correspond to every possible value of the energy @xmath58 of the system and performed a statistical estimation of its probability @xmath142 .
it is important to note that , because of the finite size of the random sample of states that we have generated ( @xmath143 configurations ) , we do not obtain states with an _ a priori _ small but non - zero probability .
this means that there is a certain number of possible values of @xmath58 for which we wrongly estimate zero probability . as the logarithmic prescription ( [ eq : afinitatlog ] )
discussed in [ subsec : discrete ] would give problems with infinities at the neighbouring values , we have only used the second prescription given by eq .
( [ eq : afinitatdisc ] ) . even in this second case , in order to avoid problems with the computation of the variance of @xmath62 ( for which a @xmath144 survives in the denominator ) , the states with an estimated zero probability have been excluded : it is as if they could not be reached by the system .
figures [ fig7 ] and [ fig8 ] show the results obtained for simulations of a @xmath145 and a @xmath146 universe , for various sizes of the system .
these are qualitatively compatible with those obtained for the system without interactions in sec .
[ subsec : example1 ] .
the uncertainty relations are fulfilled except in the very extreme case in which the number of available states of the system and/or reservoir is so small that the boundary effects are dominant .
we have shown that uncertainty relations can be established in statistical mechanics when considering a universe ( isolated from the outer world ) separated into two parts , called the system and reservoir , which are in thermodynamic equilibrium and interact by exchanging `` mechanical properties '' such as energy , volume , particles , ... if @xmath42 is one of these properties corresponding to the system , we have found that the following inequality must be satisfied , @xmath147 . here , @xmath26 is an affinity conjugated to the variable @xmath42 which determines the departure from strict equilibrium of actual microscopic configurations of the two interacting systems and which satisfies @xmath148 .
furthermore , when the entropy satisfies separability , these affinities simply reduce to the difference @xmath149 of the intensive variable conjugated to @xmath42 , in the reservoir and the system . assuming that the intensive variable of the reservoir does not fluctuate as expected for very large reservoirs
, the proposed inequality reduces to the more standard uncertainty relation involving only system variables @xmath150 furthermore , assuming that the fluctuations in the system are small , by expanding to first order , the uncertainty relation leads to the following equality : @xmath151 it is interesting to point out , than within this first - order approximation , when the relevant conjugated variables are the energy and the inverse temperature , combining eqs .
( [ eq : lindhart ] ) and ( [ eq : lindhart4 ] ) , one gets : @xmath152 where @xmath73 and @xmath74 are the heat capacities of the system and reservoir . therefore , in this last case fluctuations of inverse temperature are not , in general , expected to be proportional to energy fluctuations @xcite . only in the limit
@xmath153 is the standard result @xmath154 recovered .
the derivation of the above results has required the assumption that the probability for the boundary states ( on the boundary of the set of mechanical variables @xmath24 ) vanishes .
this assumption , which is reasonable for universes described by continuum variables , might be incorrect for the discrete case , when universes are small or too constrained .
we have numerically studied the validity of the above results in two different mechano - statistical lattice systems : the ising paramagnet and the ising ferromagnet .
we have found that the uncertainty relations hold except in the case in which the boundary states have a high probability of occurring . in particular , from the examples in section [ sec : examples ] , we can numerically evaluate a bound for this probability .
[ fig9 ] shows the uncertainty product @xmath155 against the probability of the boundary states for the ising paramagnet ( with the two possible definitions of the affinities for discrete systems given by eqs .
( [ eq : afinitatlog ] ) and ( [ eq : afinitatdisc ] ) ) and the ising ferromagnet .
the figure clearly shows that the breakdown of the uncertainity relation may occur only when the probability of the boundary states is above 0.01 . in all the other cases ,
the uncertainty product is larger than one and , in fact , tends to 1 from above when the size of the universe increases , as can be seen in the inset of fig.[fig9 ] .
these results are relevant from an experimental point of view .
focusing on the case of the energy @xmath58 and the inverse of temperature @xmath55 as conjugated variables , we can identify the reservoir with a given body and the system with the thermometer used to measure its temperature .
commonly , the thermometer is chosen so that it has a heat capacity much smaller than the heat capacity of the body . however , for small bodies , when the heat capacity of the thermometer and the body are comparable , the results derived here must be taken into account : ( i ) the inverse temperature @xmath156 measured by the thermometer con strongly fluctuate , but still its average value must be identified with the average of the inverse temperature of the body @xmath157 ; and ( ii ) the fluctuations of the difference
@xmath64 should be related to the energy fluctuations through the proposed uncertainty relation . future extensions of this work will deal with the analysis of systems described by continuous variables , where direct counting of states is not straightforward .
the results will be presented elsewhere .
this work has received financial support from the spanish ministry of innovation and science ( project mat2010 - 15114 ) .
g.t . acknowledges a grant from the university of barcelona .
we thank a.saxena for a critical reading of the manuscript . | we have analyzed the validity of uncertainty relations between the fluctuations of thermodynamically conjugated extensive and intensive variables within the field of statistical mechanics .
analysis is presented for two particular examples of small lattice systems that are in contact with reservoirs of comparable sizes : an ising paramagnet and an ising ferromagnet .
the numerical results enable determination of the range of applicablity of the proposed relations . due to the fact that the examples correspond to systems described by discrete variables , the uncertainty relations are not valid if the probability of boundary states is too high . |
over the past decade , michelson stellar interferometry has seen some tremendous advances in applicability .
it has evolved from a primarily experimental technique towards a general astrophysical observational mode extensively used by the community for galactic and extragalactic science ( see the various contributions in these proceedings ) .
observations with milli arcseconds angular resolution are now routinely performed in the near infrared .
strategic planning of next generation michelson interferometers should aim for higher resolution , higher sensitivity and image reconstruction capabilities ( synthesis or direct imaging ) .
an alternative technique to michelson interferometry that has the potential of delivering similar scientific products is the technique of intensity interferometry ( ii ) .
it is able to attain @xmath0-arcseconds resolution and provide image synthesis capacity and can be implemented in a cost effective and straightforward instrument .
the principle of ii is based on the partial correlation of intensity fluctuations of coherent light beams measured at different points in space or in time .
the fluctuations ( also called wave - noise , although `` noise '' is a bit of a misnomer ) was a well - known phenomenon at radio wavelengths , but its nature and even reality at optical wavelengths were seriously debated in the 1950s .
hanbury brown & twiss @xcite conclusively demonstrated in a series of papers how the phenomenon is firmly rooted in both theory and experiment .
the intensity fluctuations can be interpreted in a semi - classical sense as the superposition of light waves at different frequencies producing beats .
the fluctuations can also be interpreted in a quantum mechanical sense as an effect related to so - called photon bunching @xcite . at optical wavelengths ,
the resulting fluctuations are much smaller than the shot - noise component of a ( recorded ) light beam .
given that shot - noise is random , any coherence of different light beams is solely determined by the intensity fluctuation effect .
hanbury brown et al .
@xcite measured the stellar diameter for 32 stars using a dedicated stellar intensity interferometer stationed at narrabri , australia .
their pioneering experiments constituted the first successful stellar diameter measurements after the michelson & pease experiments .
we will discuss in this contribution the potential of future cherenkov telescope arrays for a revival of intensity interferometry as a mainstream high angular resolution imaging technique in astronomy .
all modern interferometers in operation are based on the principle of michelson stellar interferometry .
it provides a measure of the spatial coherence of two light beams by using the fringe contrast ( visibility ) of the beams interference pattern .
the technique requires highly accurate metrology to correct for path differences between the beams , which is on the order of the wavelength of light itself .
active correction of the optics ( adaptive optics , tip - tilt ) are required in order to make the otherwise corrugated wave front planar .
light beams are made to interfere , and allow in principle the determination of the modulus and phase of the complex visibility . in practise
however , atmospheric limitations corrupt the phase ( the fringes wander over the detector ) , and in broad - band interferometric observation a quadratic visibility estimator ( @xmath2 ) is generally employed eliminating the corrupted phase , but at the same time introducing a bias in the estimator that needs to be taken care of , e.g. , @xcite . in spectrally resolved observations a linear visibility estimator can be employed , thanks to the achromatic nature of the atmospheric disturbances , and relative phase can be retrieved .
absolute phase information with ground based interferometers can be obtained by simultaneously observing a phase reference star @xcite , or by exploiting the principle of closure phase when working with more than two telescopes , see @xcite .
intensity interferometry provides a measure of the square of the michelson fringe visibility .
this fundamental difference is the basis for some of its technical advantages .
ii is virtually insensitive to atmospheric and instrumental instabilities as no actual waves are interfered .
the technique can thus make do with relatively coarse light collectors and long baselines .
the signals from each receiver can be electronically recorded and correlated after detection .
the correlation of the multiple signals measured by a telescope array can thus all be determined for all possible pairs , and maybe even higher order correlations can be exploited .
the coherence length is set by the length of the frequency beats ( wave group ) which is of order cm rather than nm ( depending on bandwith ) , alleviating the strong constraints on accurate delay tracking .
ii is however less sensitive in measuring coherence than michelson interferometry , as the sought fluctuations are very small relative to the shot - noise of the photon stream .
however , without stringent requirements for the light collectors , the decreased sensitivity can be offset by large coarse mirrors in order to maximise photon collection .
in addition @xcite have shown that under certain conditions higher order correlations will increase the sensitivity ( see @xcite ) .
a second drawback is no phase information is contained within the measurement .
although this does not impede the measurement of centro - symmetric objects ( indeed as was done with the narrabri stellar intensity interferometer ) , it has been shown that the phase can be recovered using correlations between more than two signals @xcite . also superimposing a coherent beam from a known reference source on
the light beam of the target source would allow the recovery of the phase @xcite , see also @xcite .
imaging air cherenkov telescope ( iact ) arrays are multi - telescope arrays designed to image air showers that are produced by high energy particles and @xmath3-ray photons ( @xmath4 gev ) that impinge on the earth s atmosphere .
the air showers consist of secondary particles some of which carry electrical charge and produce cherenkov light . as the cherenkov radiation is faint ( depending on the energy of the incident particle / photon ) , large collecting areas are required to obtain decent signal - to - noise ratios .
the flash of cherenkov radiation produced is also very brief ( a few nanoseconds ) and fast photon counting detectors are therefore mandatory .
the advantage of employing telescope arrays in air - shower observations is the ability to reconstruct the spatial geometry of the shower in the earth s atmosphere .
this allows one to make a distinction between cosmic ray showers and @xmath3-ray showers , and also to find the spatial direction of the incident particle / photon , thus obtaining angular information on the galactic or extragalactic source . in short , iacts aim to determine the energy and the astrophysical source of @xmath3-ray photons by observing the optical light of induced cherenkov radiation .
it should be clear that the basic technical specifications of an iact array as touched upon here are very similar to the ones of an intensity interferometer as briefly described in the previous section .
a more detailed description of the components of iact arrays relevant for ii ( fast photon detection , signal communication , correlators ) can be found in @xcite .
currently , there are two major and successful iact arrays operational .
the h.e.s.s .
array @xcite consists of 4 telescopes of 108m@xmath5 each .
they are arranged in a square with 120 m side length .
the veritas @xcite array is similar and also consists of 4 telescope with 110m@xmath5 light collecting area .
in contrast to hess , it has variable baselines ranging from 34 to 109 metres .
hess will be upgraded to hess ii with one extra 30 m dish telescope placed at the centre of the configuration , its completion foreseen for the 4th quarter of 2009 .
future iact projects are currently under study , such as agis[multiblock footnote omitted ] and cta[multiblock footnote omitted ] , and aim at increasing the number of telescopes up to 100 , each with a @xmath6m@xmath5 light collectors arranged over an area covering 1km@xmath5 or more . these envisioned projects ( expected to be operational around 2015 ) would not only benefit the resolving power and sensitivity for the high - energy science purposes of iacts , but they are in harmony with the requirements of a more sensitive intensity interferometer with a higher angular resolution and denser ( u , v)-coverage @xcite .
future iacts will offer thousands of baselines from 50 m to at least 1 km and when used as an intensity interferometer may attain angular resolution down to a few tens of @xmath0-arcseconds . _
left : _ proposed lay - out for the future cta .
small red dots are the 85 @xmath7 dishes , large blue dots are the four @xmath8 dishes ( adapted from @xcite ) . _
sensitivity estimate as function of the 48 non - redundant baselines for a 5@xmath9 detection , in a 5 hr integration on a centro - symmetric object with 50% visibility .
final values depend on signal bandwith and cta design details , see @xcite.,title="fig:",width=226,height=226 ] _ left : _ proposed lay - out for the future cta .
small red dots are the 85 @xmath7 dishes , large blue dots are the four @xmath8 dishes ( adapted from @xcite ) . _
sensitivity estimate as function of the 48 non - redundant baselines for a 5@xmath9 detection , in a 5 hr integration on a centro - symmetric object with 50% visibility .
final values depend on signal bandwith and cta design details , see @xcite.,title="fig:",width=302,height=226 ]
limiting @xmath10 of the cta concept is illustrated in the right panel of fig.1 with simulated performance that meet the goal sensitivity for @xmath3-ray astronomy @xcite .
targets are limited to a @xmath11 for a s / n = 5 , and a 5 hours integration in case of 50% visibility ( see @xcite ) .
these specifications allow important interferometry studies regarding binary stars , stellar radii and pulsating stars with unprecedented resolution on @xmath0-arcsecond scales .
we highlight three potential science cases below .
[ [ star - formation ] ] * star formation * + + + + + + + + + + + + + + + + key questions relating to the physics of mass accretion and pre - main sequence ( pms ) evolution can be addressed by means of very high resolution imaging as provided by the next generation iacts and ii .
they involve the absolute calibration of pms tracks , the mass accretion process , continuum emission variability , and stellar magnetic activity .
the technique may allow us to the resolve spot features on the stellar surface .
hot spots deliver direct information regarding the accretion of material onto the stellar surface .
cool spots , on the other hand , may cover 50% of the stellar surface , and they are the product of the slowly decaying rapid rotation of young stars
. imaging them will constrain ideas regarding the interplay of rotation , convection , and chromospheric activity as traced by cool spots .
it may provide direct practical application as the explanation for the anomalous photometry observed in young stars @xcite . in practise ,
about 50 young stars with @xmath12 are within reach of future iacts . in the last decade
several young coeval stellar groups have been discovered in close proximity ( @xmath1350pc ) to the sun .
famous examples are the tw hydra and @xmath14 pic comoving groups . the majority of the spectral types within reach range between a and g - type .
their ages lie within the range 8 to 50 myr ( see for an overview ) .
the age intervals ensures that a substantial fraction of the low - mass members are still in their pms contraction phase .
measurement of their angular size can be used in the calibration of evolutionary tracks , fundamental in deriving the properties of star forming regions and young stellar clusters .
the proximity of the comoving groups ensures that their members are bright .
their proximity renders the comoving group also relatively sparse making them very suitable , unconfused targets despite the large optical psf of a few arcminutes .
the sparseness is also the reason for incomplete group memberships , making it likely that the number of young stars close to the sun will increase with the years to come .
[ [ distance - scale - and - pulsating - stars ] ] * distance scale and pulsating stars * + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + measuring diameters of cepheids is a basic method with far reaching implications .
a radius estimate of a cepheid can be obtained using the baade - wesselink method .
the baade - wesselink method relies on the measurement of the ratio of the star size at times @xmath15 and @xmath16 , based on the luminosity and colour .
combined with a simultaneous measurement of the radial velocity , this method delivers the difference in the radius between @xmath15 and @xmath16 . with the known difference and ratio of the radius at two times
, one can derive the radius of the cepheid . combining ii angular size measurement with the radius estimate one
obtains the distance to the cepheid ( see @xcite ) .
this makes possible the calibration of the all important cepheid period - luminosity relation using local cepheids .
a count of cepheids observed with _
hipparcos _ shows that at least 60 cepheids with @xmath17 are in reach with future iacts .
[ [ rapidly - rotating - stars ] ] * rapidly rotating stars * + + + + + + + + + + + + + + + + + + + + + + + + as a group , classical be stars are particularly well - known for their close to break - up rotational velocities as deduced from photospheric absorption lines . in addition the stars show balmer line emission firmly established as due to gaseous circumstellar disks , that appear and disappear on timescales of months to years .
these two properties are somehow related , but many open questions regarding the detailed physical processes at play exist .
the be - phenomenon is an important phenomenon given the number of stars and stellar physics involved ( fraction of be stars to normal b - type peaks at nearly 50% for b0 stars , ) .
absorption lines will however never provide the final answer to their actual rotational velocity due to strong gravity darkening at the equator and brightening at the pole areas .
direct measurement of the shape of the rotating star is not hampered by gravity darkening , and provides a direct indication of the rotational speed ( see , e.g. , @xmath18 eri with the vlti , ) . the be star disk formation and dissolution activity is little understood .
photometric observations of be star disks seem to indicate that they may actually evolve into ring structures before disappearing into the interstellar medium ( e.g. ) .
the disk s bremsstrahlung can constitute @xmath19 of the total light in @xmath20-band .
there are about 300 be stars[multiblock footnote omitted ] brighter than @xmath21 , roughly corresponding to a distance limit of 700pc . signifying that be star phenomena can be probed in depth with iact based ii .
technical requirements for a relatively sensitive and @xmath0-arcsecond resolution synthesis intensity interferometer go hand in hand with the designs for next generation air cherenkov telescope arrays .
this prompts close study of the possibilities in designing and building a iact instrument that incorporates the ii capability . in principle
there is no competition between the two modes , as @xmath3-ray observations need to be executed when the moon is less than half full .
this leaves half the available night time of an iact for interferometric observations .
the renewed interest in ii has resulted in the formation of an iau working group on intensity interferometry .
laboratory experiments are performed to test and demonstrate the various aspects of ii integrated in a iact .
a pair of 3-m telescopes is now available in utah , usa ( star base utah 2008[multiblock footnote omitted ] ) for testing the techniques in a realistic astronomical environment . in conclusion ,
the time is now to assess the applicability of ii as a future high resolution imaging mode that allows science to enter the realm of astrophysics on @xmath0-arcsecond angular scales .
this contribution is partly based on two presentations @xcite given at the meeting `` the universe at sub - second timescales '' held in edinburgh , september 2007 .
10 url # 1#1urlprefix[2][]#2 r and twiss r q 1956 _ nature _ * 178 * 1046 s , daniel m , de wit w j , hinton j a , jose e , holder j a , smith j and white r j 2008 _ high time resolution astrophysics : the universe at sub - second timescales _
( _ american institute of physics conference series _ vol 984 ) p 205 w j , le bohec s , hinton j a , white r j , daniel m k and holder j 2008 _ high time resolution astrophysics : the universe at sub - second timescales _ ( _ american institute of physics conference series _ vol 984 ) p 268 | in this poster contribution we highlight the equivalence between an imaging air cherenkov telescope ( iact ) array and an intensity interferometer for a range of technical requirements .
we touch on the differences between a michelson and an intensity interferometer and give a brief overview of the current iact arrays , their upgrades and next generation concepts ( cta , agis , completion 2015 ) .
the latter are foreseen to include 30 - 90 telescopes that will provide 400 - 4000 different baselines that range in length between 50 m and a kilometre .
intensity interferometry with such arrays of telescopes attains 50@xmath0-arcseconds resolution for a limiting @xmath1 .
this technique opens the possibility of a wide range of studies , amongst others , probing the stellar surface activity and the dynamic au scale circumstellar environment of stars in various crucial evolutionary stages .
here we discuss possibilities for using iact arrays as optical intensity interferometers . |
quantum computers could solve some problems faster than classical computers @xcite . performing a quantum computation relies on the ability to preserve the coherence of quantum states long enough for gates composing the algorithm to be implemented . in practice ,
the quantum coherence is sensitive to the uncontrolled environment and easily damaged by the interactions with the environment , a process called decoherence @xcite . to protect the fragile quantum coherence needed for quantum computation , schemes of quantum error correction ( qec ) and fault- tolerant quantum computation
have been developed @xcite .
the 3-bit qec code was implemented in a liquid - state nmr quantum information processor in 1998 as the first experimental demonstration of qec @xcite .
more recently , it has been implemented in trapped ion and solid - state systems @xcite . here
we report on using the grape algorithm @xcite to implement a high fidelity version of the 3-bit qec code for phase errors in liquid state nmr .
the errors due to natural transversal relaxation are shown to be suppressed to a first order . in comparison with the work performed in 1998 @xcite ,
the pulse sequence fidelity is improved by about @xmath2 , and the reduction of the first order in the decay of the remaining polarization after error correction is improved by a factor of @xmath3 .
the advantage of the qec is obtained although the extra operations for protecting the quantum states in qec are subject to errors in implementation .
in the current implementation , we use @xmath4c labelled trichloroethylene ( tce ) dissolved in d - chloroform as the sample .
data were taken with a bruker drx 700 mhz spectrometer .
the structure of the molecule and the parameters of the spin qubits are shown in fig . [ figmol ] , where we denote h as qubit 1 , c@xmath5 as qubit 2 and c@xmath6 as qubit 3 .
the hamiltonian of the three - spin system can be written as @xmath7\ ] ] where @xmath8 , @xmath9 , @xmath10 denote the pauli matrices with @xmath11 indicating the spin location , @xmath12 denotes the chemical shift of spin @xmath11 , and @xmath13 denotes the spin - coupling between spins @xmath11 and @xmath14 .
the two carbon spins are treated in the strongly coupled regime , because the difference in frequencies between the two carbons is not large enough for the weak coupling approximation @xcite .
we exploit radio - frequency ( r.f . )
spin selection techniques to improve the linewidth , and hence the coherence , of the ensemble qubits @xcite . the effect of pulse imperfections due to r.f .
inhomogeneities is reduced by spatially selecting molecules from a small region in the sample through the r.f . power .
we choose c@xmath15 as the qubit to carry the state for encoding and the output state after decoding and error correction .
the labelled pseudo - pure states @xmath16 and @xmath17 , used as the reference states with blank ancilla , are prepared by the circuit in ref .
@xcite , where the order is arranged as qubits @xmath18 to @xmath19 and @xmath20 .
the qubit readout is performed on c@xmath5 , and the signals are normalized with respect to @xmath16 or @xmath17 , for different input states .
the quantum network used for implementing the qec code is shown as fig .
[ figcir ] ( a ) , where @xmath21 is chosen as @xmath22 , @xmath23 and @xmath24 , in separate sequences .
we optimize the encoding operation , and the decoding operation combined with the error correction as two grape pulses @xcite with theoretical fidelity @xmath25 . to test the ability of the code to correct for the natural dephasing errors due to the transversal relaxation of the spins , the internal spin hamiltonian ( [ ham ] )
is refocused during the time delay implemented between the encoding and decoding processes .
the refocusing pulse sequence is shown in fig .
[ figcir ] ( b ) where the selective @xmath26 pulses applied to spin h are hard rectangle pulses with a duration of @xmath27 @xmath28s , while the @xmath26 pulses applied to c@xmath15 or c@xmath29 are grape pulses with a duration of @xmath30 ms .
taking into account the strong coupling in the hamiltonian ( [ ham ] ) , we choose the phases of the @xmath26 pulses shown in fig .
[ figcir ] ( b ) to obtain a fidelity @xmath31 , where @xmath32 denotes the simulated unitary implemented by the pulse sequence , and @xmath33 denotes the identity operation .
we choose the input states as @xmath34 , @xmath23 and @xmath24 , and measure the polarization that remains after error correction in @xmath35 .
the polarization ratios are denoted as @xmath36 , @xmath37 and @xmath38 .
we use entanglement fidelity " , represented as @xmath39 to characterize how well the quantum information in @xmath21 is preserved @xcite .
the experimental results for qec are shown in fig .
[ figres ] ( a ) . for each delay time
, five experiments are repeated in order to average the random experimental errors in implementation .
the results of error correction ( ec ) are represented by @xmath40 . by averaging the points for each delay time ,
we obtain the averaged entanglement fidelity @xmath41 shown as @xmath42 , which can be fitted to @xmath43 , with relative fitting error @xmath44 , shown as the thick dash - dotted curve . in order to estimate the performance of the error correction for the encoded states , we calculate the entanglement fidelity of decoding ( de ) through measuring the remaining polarization before the application of the toffoli gate , used as the error - correcting step . in this case , the decoding operation is implemented by one grape pulse with theoretical fidelity @xmath25 .
similar to the measurement for error correction , we also repeat five experiments for each delay time .
the results are shown as @xmath45 in fig .
[ figres ] ( a ) , and the data points after average are marked by + , which can be fitted as @xmath46 , with relative fitting error @xmath47 , shown as the thick solid curve . here
the ratio of the first order decay terms for the two fits is found to be @xmath48 .
the important reduction of the first order decay term indicates the high quality of state stabilization provided by qec . as a comparison
, we include the experimental data from ref .
@xcite , which are marked as @xmath49 and @xmath50 in fig .
[ figres ] ( b ) for the results of qec and decoding .
the data can be fitted as @xmath51 and @xmath52 with relative fitting errors @xmath53 and @xmath54 , respectively .
the ratio of the first - order decay terms is @xmath55 . in implementing the qec code ,
the operations associated with encoding , decoding and error correction are subject to errors , which would lower the ability of the code to protect the quantum states .
to estimate the effects of the errors , we measure the free evolution decay ( fed ) of @xmath21 under the refocusing sequence shown in fig .
[ figcir ] ( b ) .
five experiments are repeated for each delay time , and the experimental data for @xmath41 are shown as @xmath56 in fig .
[ figres ] ( a ) .
the averaging points , shown as @xmath57 , can be fitted as @xmath58 with relative fitting error @xmath59 , shown as the dashed curve .
the ratio of the first order decay terms in the fits of fed and ec is @xmath60 . through comparing the results of qec and fed
, one can find that the errors removed by the qec code can exceed the errors introduced by the extra operations required by the code for delay time @xmath61 s ( @xmath62 of c@xmath5 s @xmath63 ) .
the pulse durations for encoding , decoding , and the combination of decoding and error correction are 8 ms , 8ms , and 13.6 ms , respectively .
we exploit the results from simulation with ideal pulses to estimate the errors due to the imperfection in pulse implementation . in simulation , we choose an uncorrelated error model for @xmath64 errors and ignore @xmath65 errors @xcite .
we represent the measured fidelity as @xmath66 , where @xmath67 denotes the ideal fidelity by simulation and @xmath68 denotes a factor to estimate the deviation between experiment and simulation .
one should note that the theoretical entanglement fidelity of de is the same as fed @xcite . by fitting the data , we obtain @xmath69 , @xmath70 and @xmath71 for ec , de and fed , respectively .
the fitting results are shown as the thin dash - dotted , solid and dashed curves in fig .
[ figres ] ( a ) . from the simulation results , we estimate the errors in implementing the operations associated with the qec codes are about @xmath72 for de and @xmath73 for ec .
we optimize the encoding , decoding , and error correction as grape pulses with high theoretical fidelities ( @xmath25 ) .
the refocusing sequence is exploited to suspend the evolution of the hamiltonian ( [ ham ] ) with high fidelity ( @xmath74 ) .
the quality of readout signals are further improved by r.f . selection . compared with the experimental results of qec obtained in 1998 @xcite ,
the pulse sequence fidelity is improved by about @xmath2 . by the comparison with the free evolution decay
, one can benefit from qec even when errors exist in implementing the operations required for qec .
the improvement provided by the error correction is also demonstrated by the reduction of the first order in the decay of the remaining polarization after error correction , compared with the decay of the encoded states recovered by decoding and free evolution decay of the input states .
the `` qec advantage '' for the encoded states is improved by a factor of @xmath3 from the 1998 result . in the current experiment , the second order term in the decay after
error correction is larger than the previous experiment , because of the larger phase errors due to the shorter @xmath63 time constants [ see fig .
[ figmol ] ( b ) ] , noting that @xmath63 s are @xmath19 s for h , @xmath75 s for c@xmath15 and @xmath76 s for c@xmath29 in the previous experiment @xcite .
the experimental errors arise mainly from the imperfection in implementing the grape pulses .
additionally inhomogeneities of magnetic fields and the limitation of @xmath65 also contribute to errors .
the authors acknowledge professor d. g. cory for helpful discussions .
+ @xmath77 current address : department of physics , massachusetts institute of technology , cambridge , ma , usa 02139 e. knill , r. laflamme and w. h. zurek , science , * 279 * , 342 ( 1998 ) ; a. yu .
kitaev , russ .
surv . * 52 * , 1191 - 1249 ( 1997 ) ; d. aharonov and m. ben - or , _ proceedings of the 29th annual acm symposium on theory of computing _ , 176 ( acm press , 1997 ) ; e. knill , nature , * 434 * , 39 ( 2005 ) ; d. gottesman , _ encyclopedia of mathematical physics _ , edited by j .-
francoise , g. l. naber , and s. t. tsou , 196 ( elsevier , oxford , 2006 ) .
n. khaneja , t. reiss , c. kehlet , t. schulte - herbruggen and s. j. glaser , j. magn . reson . * 172 * , 296 ( 2005 ) ; c. a. ryan , c. negrevergne , m. laforest , e. knill and r. laflamme , phys .
a * 78 * , 012328 ( 2008 ) . and c@xmath29 .
( b ) the relaxation times @xmath78 s are measured by the standard inversion recovery sequence .
@xmath63 s are measured by the hahn - echo with one refocusing pulse , by ignoring the strong coupling in the hamiltonian ( [ ham]).,width=384 ] noise is introduced by a variable time delay implemented by the pulse sequence ( b ) which refocuses the evolution of the hamiltonian ( [ ham ] ) to an identity operation with theoretical fidelity higher than @xmath79 . in ( b )
the unfilled rectangle represents a hard @xmath26 pulse with duration of @xmath27 @xmath28s .
the filled rectangle represents a grape @xmath26 pulse selective for c@xmath15 or c@xmath29 with duration of @xmath30 ms .
the phases of the pulses are denoted above the rectangles.,width=384 ] for ec , @xmath45 for de and @xmath56 for fed .
the averages are shown as @xmath42 , + and @xmath57 , which can be fitted as @xmath80 with relative fitting error @xmath44 , @xmath46 with relative fitting error @xmath47 and @xmath58 with relative fitting error @xmath59 , shown as the thick dash - dotted , solid and dashed curves , respectively .
the ratios of the first - order decay terms in the fitted curves are calculated as @xmath48 for de and ec , and @xmath60 for fed and ec , respectively .
the thin dash - dotted , solid and dashed curves show the fitting results using the ideal data points from simulation by introducing factors of @xmath81 , @xmath70 and @xmath71 for ec , de and fed , respectively .
( b ) results in the previous experiment @xcite , shown as the data marked by
@xmath49 `` and @xmath50 '' for ec and de , which can be fitted as @xmath82 and @xmath83 with relative fitting errors @xmath53 and @xmath54 , respectively .
the ratio of the first - order decay terms is @xmath55.,width=576 ] | more than ten years ago a first step towards quantum error correction ( qec ) was implemented [ phys .
rev .
lett .
81 , 2152 ( 1998 ) ] .
the work showed there was sufficient control in nuclear magnetic resonance ( nmr ) to implement qec , and demonstrated that the error rate changed from @xmath0 to approximatively @xmath1 . in the current work
we reproduce a similar experiment using control techniques that have been since developed , such as grape pulses .
we show that the fidelity of the qec gate sequence , and the comparative advantage of qec are appreciably improved .
this advantage is maintained despite the errors introduced by the additional operations needed to protect the quantum states . |
the first compound light microscopes constructed in the 16th and 17th centuries enabled scientists to inspect matter and biological specimens at the microscopic level @xcite . in 1873 , ernst abbe formulated a fundamental limit for the resolution of an optical imaging system based on the diffraction theory of light @xcite .
at the same time the fabrication and development of microscopes and lenses were transformed from empirical optimizations to schemes based on quantitative calculations and theoretical considerations . in the 20th century various contrast modalities were developed that allow one to detect very small signals and to measure characteristic properties of a specimen with high specificity . finally , during the last two decades several revolutionary methods were conceived and experimentally demonstrated , which substantially enhanced the optical resolution down to the nanometer scale ( shown in fig .
[ fig_opticalresolution ] ) .
the awarding of the 2014 nobel prize in chemistry to eric betzig , stefan hell and william e. moerner for their pioneering work in `` super - resolution '' fluorescence microscopy corroborates its promise for many advanced investigations in physics , chemistry , materials science and life sciences .
fluorescence microscopy down to the single molecule level has been reviewed in many recent articles and books @xcite . despite the immense success of fluorescence microscopy
, this technique has several fundamental shortcomings . as a result
, many ongoing efforts aim to conceive alternative modes of microscopy based on other contrast mechanisms .
furthermore , having overcome the dogma of the resolution limit , scientists now focus on other important factors such as phototoxicity and compatibility with live imaging , higher speed , multiscale imaging and correlative microscopy . in this review article
, we present a concise account of some of the current trends and challenges .
every measurement needs an observable , i.e. a signal . in the case of optical microscopy ,
one correlates a certain optical signal from the sample with the spatial location of the signal source .
scattering is the fundamental origin of the most common signal or contrast mechanism in imaging . indeed ,
when one images a piece of stone with our eyes we see the light that is scattered by it although in common language one might speak of reflection .
the scattering interaction also leads to a shadow in transmission ( see fig .
[ fig_transrefl ] ) . in conventional microscopy , one speaks of trans - illumination if one detects the light transmitted through the sample and epi - illumination if one detects the signal in reflection . already the pioneers of early microscopy experimented with different types of illumination to generate stronger contrasts .
even today , a good deal of instrumentation in common microscopes focuses on the illumination path .
for example , in a particularly interesting scheme one adjusts the illumination angle such that a negligible amount of it is captured by the finite solid angle of the detection optics . in this so - called dark - field microscopy
, one emphasizes parts of the sample that scatter light in an isotropic fashion .
such oblique illumination was already exploited in the early days of microscopy @xcite . during the past century
, various methods were developed to improve the contrast in standard microscopy .
for example , polarization microscopy techniques can be used to examine the anisotropy of birefringent materials such as minerals @xcite .
some of the most important contrast mechanisms exploit the spectroscopic information of the sample and thus introduce a certain degree of specificity .
the prominent example of these is fluorescence microscopy , where fluorophores of different absorption and emission wavelengths are employed to label various parts of a biological species @xcite . over the years , the developments of fluorescence labeling techniques such as immunofluorescence @xcite , engineered organic fluorescent molecules @xcite and fluorescent proteins @xcite have continuously fueled this area of activity .
however , fluorescence labeling has many disadvantages such as photobleaching and most importantly the need for labeling itself . to address some of the limitations of standard fluorescence imaging
, scientists have investigated a range of multiphoton fluorescence methods such as two - photon absorption @xcite .
the strong dependence of multiphoton excitation processes on intensity allows one to excite a small volume of the sample selectively only around the focus of the excitation beam .
this leads to minimizing the fluorescence background and photobleaching . aside from its capacity for optical sectioning , this technique makes it possible to perform tissue imaging because the long - wavelength excitation light penetrates deeper into the biological tissue .
the ultimate freedom from fluorescence markers is the use of label - free contrast mechanisms .
for example , raman microscopy generates contrast through the inelastic scattering of light that is selective on the vibrational and rotational modes of the sample molecules @xcite and is , thus , very specific to a certain molecule in the specimen .
the main difficulty of raman microscopy lies in its extremely weak cross section .
methods such as coherent anti - stokes raman scattering ( cars ) @xcite or stimulated raman scattering @xcite improve on the sensitivity to some extent although they remain limited well below the single molecule detection level .
some other interesting label - free contrasts are based on the harmonic generation of the illumination or four - wave mixing processes through the nonlinear response of the sample @xcite .
for instance , it has been shown that collagen can be nicely imaged through second harmonic generation ( shg ) @xcite . to conclude this section
, we return to the most elementary contrast mechanisms , namely transmission and reflection .
the fundamental underlying process in these measurements is interference .
consider again the situation of fig .
[ fig_transrefl ] , whereby now we reduce the size of the object to a subwavelength scatterer such as a nanoparticle ( shown in fig .
[ fig_iscat]a ) .
a laser beam with power @xmath0 illuminates the nanoparticle of diameter @xmath1 lying on a glass substrate .
a fraction of the incoming power @xmath2 is scattered by the object and another part of it serves as a reference @xmath3 , which can either be the transmitted light or the light that is reflected from the glass interface .
the scattered and the reference components of the light reach the detector and interfere if their path difference is much smaller than the coherence length of the illumination beam .
in the case of a reflection configuration with field reflectivity @xmath4 , the detected power @xmath5 on the detector is given by @xmath6 where @xmath7 signifies the scattering cross - section of the particle and @xmath8 denotes a phase term that includes the gouy phase @xcite .
interferometric scattering contrast was explicitly formulated in the context of detection of nanoscopic particles in our laboratory @xcite and has been coined iscat @xcite .
the central process is , however , quite general and related to the measurement of extinction by an object on a beam of light . according to the optical theorem ,
the extinction signal can stem from scattering loss or absorption @xcite .
we note that although the interaction of light with objects that are larger than a wavelength is expressed in terms of reflection instead of scattering , the signal on the detector can still be written in the form of eq .
( [ interference ] ) .
having understood the underlying interferometric nature of an extinction measurement , it is now clear that one can also perform the measurement by manipulating the reference beam in eq .
( [ interference ] ) .
an early example of such a variation was demonstrated by zernike in the context of phase contrast microscopy @xcite . here a part of the illumination is phase shifted and then mixed with the light that is transmitted through the sample .
another related technique is known as differential interference contrast microscopy ( dic ) put forth by nomarski @xcite . a very similar method
to dic that is somewhat simpler to implement and therefore very common in commercial microscopes is hoffman modulation contrast @xcite .
although simple transmission , reflection , phase contrast and dic share the same fundamental contrast as that of iscat , only the latter has tried to address the issue of sensitivity and its extension to very small nanoparticles and single molecules .
this brings us to the general question of sensitivity . for any given contrast mechanism
, one can ask `` how small of an object can one detect ? '' .
in other words , what is the sensitivity . to have high sensitivity , one needs a sufficiently large signal from the object of desire , and the trouble is that usually the signal diminishes quickly as the size of the object is reduced .
so , one needs to collect the signal efficiently and employ very sensitive detectors .
the detector can either be a point detector as it is most often used in scanning confocal techniques or a camera in the case of wide - field imaging .
the performance of a light detector can be generally described by its quantum efficiency , the available dynamic range and its time resolution @xcite . for the longest time in the history of microscopy
the human eye was the only available detector .
with its detection threshold of only a few photons it was a better detector than photographic plates and films even long after these became available @xcite .
photon counting detectors had emerged by the 1970s , starting with photo - multiplier tubes ( pmt ) @xcite and later followed by semiconductor devices that are able to detect single photons .
these single - photon avalanche diodes ( spad ) can nowadays achieve quantum efficiencies above 50 % with timing resolutions on the order of tens of picoseconds .
in the 1990s , cameras like the charge - coupled device ( ccd ) and fast active pixel sensors using complementary metal - oxide - semiconductor ( cmos ) technology reached the single - photon sensitivity with high quantum efficiencies .
today , the best ccd cameras using electron multiplication can achieve quantum efficiencies better than 95 % in the visible part of the spectrum with a readout noise of a fraction of a photo - electron .
is illuminated by a light beam with power @xmath0 .
part of the light ( @xmath3 ) will serve as a reference beam and part of the light is scattered by the particle ( @xmath2 ) . in the imaging plane
the two beams can interfere leading to an enhancement of the scattering signal . *
( b ) * examples of differential iscat images are shown for two protein types .
individual molecules are marked with arrows .
scale bars : @xmath9 .
( b : modified with permission from @xcite . ) ] an additional important issue regarding detection sensitivity concerns the background signal , which can swamp the signal .
if the detector dynamic range ( the ratio of the highest and the smallest possible amount of light that can be measured ) is large enough , one can subtract the background .
however , temporal fluctuations on the background usually limit this procedure in practice . in general
, the image sensitivity can be quantified in terms of the signal - to - noise ratio ( snr ) , @xmath10 where @xmath11 denotes the mean and @xmath12 the standard deviation .
the sensitivity of fluorescence microscopy was taken to the single - molecule limit towards the end of the early 1990s @xcite .
the factors leading to the success of this field were 1 ) access to detectors capable of single - photon counting , 2 ) the ability to suppress the background by using spectral filters , 3 ) preparation of clean samples .
although single - molecule fluorescence microscopy has enabled a series of spectacular studies in biophysics , fluorescence blinking and bleaching as well as low fluorescence quantum yield of most fluorophores pose severe limits on the universal applicability of this technique .
the low cross sections of raman and multiphoton microscopy methods also hinder the sensitivity of these methods . only in isolated cases , where local field enhancement has been employed ,
have been reports of single - molecule sensitivity @xcite . for a long time , single - molecule sensitivity in extinction was also believed not to be within reach because the extinction contrast of a single molecule is of the order of @xmath13 , making it very challenging to decipher the signal on top of laser intensity fluctuations .
nevertheless , various small objects have been detected and imaged using iscat , including very small metallic nanoparticles @xcite , single unlabeled virus particles @xcite , quantum dots even after photobleaching @xcite , single molecules @xcite and even single unlabeled proteins down to a size of 60 kda @xcite ( shown in fig .
[ fig_iscat]b ) . here
, it is important to note that the power that is scattered by a rayleigh particle ( @xmath2 ) follows a @xmath14 law where @xmath1 denotes the particle diameter . the interference term @xmath15 ( cf .
. [ interference ] ) contains the scattering field or in other words the polarizability of the particle , which is proportional to @xmath16 .
therefore , this term will dominate the detected power for very small objects .
the high sensitivity of iscat means , however , that any slight variation in the index of refraction or topography can lead to a sizable contrast .
hence , it is important to account for fluctuations of the index of refraction , length or absorption in the sample .
one of the most immediate functions that the layperson associates with a microscope is its ability to reveal the small features of a sample .
the principle of operation of a microscope is typically described using ray optics . however ,
when dimensions are to be investigated of the order of the wavelength of visible light , i.e. 400 - 800 nm , we must consider the wave properties of light such as interference and diffraction @xcite .
therefore one can not achieve an arbitrarily high resolution simply by increasing the magnification of the lens arrangement .
ernst abbe pioneered a quantitative analysis of the resolution limit of an optical microscope @xcite .
he considered imaging a diffraction grating under illumination by coherent light @xcite .
abbe argued that one would recognize the grating if one could detect at least the first diffraction order ( see fig .
[ fig_psf]a ) . in the case of an immersion microscope objective with circular aperture and direct on - axis illumination ,
the abbe diffraction limit of resolution @xmath1 reads @xmath17 where @xmath18 is the wavelength of light and @xmath19 was introduced as the numerical aperture ( illustrated in fig . [
fig_psf]b ) .
here @xmath20 is the refractive index of the medium , in which the microscope objective is placed , and @xmath21 denotes the half - angle of the light cone that can enter the microscope objective .
air has a refractive index of about @xmath22 limiting the @xmath23 of dry microscope objectives to less than unity . by filling the space between cover glass and an immersion microscope objective with a high index material
, the numerical aperture can be increased . already in the original publication ,
ernst abbe discussed how this diffraction - limited resolution can be improved if the illumination comes at an angle with respect to the optical axis , making it possible to collect higher diffraction orders ( cf .
[ fig_psf]a ) . in this case
, the diffraction limit is determined by the sum of the numerical apertures of the illumination lens and the collection lens .
if the angles of incidence and collection are identical , a factor of @xmath24 is obtained , leading to the famous abbe formula @xmath25 considering that the optical response of an arbitrary object can be fourier decomposed , abbe s formula can be used as a general criterion for resolving its spatial features . at about the same time
, hermann von helmholtz developed a more elaborate mathematical treatment that he published one year after abbe @xcite .
helmholtz also discussed incoherent illumination and showed that eq .
( [ eqn_incoherent ] ) holds in that case too . in a nutshell , for an incoherent illumination
, the light is scattered from the individual nanoscopic parts of the structure rather than diffracted in a well - defined order .
thus one can consider the illumination to be from every direction @xcite .
lord rayleigh was the first to consider self - luminous objects , which also emit incoherently @xcite .
additionally , he discussed different types of objects , different aperture shapes and the similarities in the diffraction limit for microscopes and telescopes .
although abbe s resolution criterion is more rigorous , a more commonly known formulation of the resolution for spectrometers and imaging instruments is the rayleigh criterion , @xmath26 it states that two close - lying points are considered resolved if the first intensity maximum of one diffraction pattern described by an airy disc @xcite coincides with the first intensity minimum of the other diffraction pattern @xcite .
equation ( [ eqn_rayleigh ] ) is somewhat arbitrary but it comes very close to abbe s criterion .
rayleigh based his definition upon the physiological property of the human eye , which can only distinguish two points of a certain intensity difference . in practice
, the full - width at half - maximum ( fwhm ) of the point - spread function ( psf ) also provides a useful criterion for the resolution of a microscope because two overlapping psfs that are much closer than their widths can no longer be resolved . for an immersion oil microscope objective with @xmath27 operating at a wavelength of 500 nm , the psf is about 220 nm ( see fig . [ fig_psf]b ) .
the axial width of the psf is about 2 - 3 times larger , in this case amounting to about 400 nm . nm and a numerical aperture na = 1.49 .
the lateral and axial fwhms of the psf amount to about 220 nm and 400 nm , respectively . ]
it is worth mentioning that the psf can also take different forms depending on the employed optical beam .
for example , it has been shown that the psf of a focused doughnut beam with radial polarization can be made somewhat smaller than that of a conventional tem@xmath28 mode @xcite .
the origin of this effect is the vectorial character of optical beams .
fluorescence is one of the most important contrast mechanisms because it offers the possibility of specific labeling .
however , since the spontaneous emission of a fluorophore does not preserve the coherence of the illumination , the signal is incoherent . as we shall see
shortly , one can engineer the illumination to increase the resolution by a factor of @xmath24 .
one of the possibilities for improving the resolution is confocal microscopy .
although the principle was patented by marvin minsky in 1957 , it took 20 years until the invention of suitable lasers and the progress in computer - controlled data acquisition opened the door for its widespread use @xcite .
in contrast to conventional wide - field illumination , where the full field of view is illuminated and imaged onto a camera , scanning confocal microscopy uses spatial filtering to ensure that only a diffraction - limited focal volume is illuminated and that only light from this focal volume can reach the detector .
an image is then produced by raster scanning this confocal volume across the sample . since out - of -
focus light is effectively suppressed , the method allows for higher contrast and offers the ability to perform optical sectioning to acquire 3d images .
the lateral size of the psf can be improved by a factor of @xmath29 in confocal microscopy @xcite . however , in reality this factor depends on the coherence properties of the imaging light and the finite size of the detection pinhole @xcite .
the latter is usually set to a value about the size of the point spread function so as to not lose any signal . as a result
, only a marginal improvement of the resolution can be achieved in confocal microscopy .
a particularly interesting mode of confocal microscopy is image scanning microscopy , where an image is recorded on a camera at each scan point @xcite .
it has been shown that one can computationally reconstruct an image with a resolution that is improved by up to a factor of @xmath24 from the resulting image stack .
the scanning feature of confocal microscopy limits its speed and , thus , wide - field approaches are generally favored .
an attractive and powerful strategy to improve the lateral resolution of wide - field fluorescence microscopy by up to a factor of @xmath24 is offered by structured - illumination microscopy ( sim ) @xcite . here
, the sample is illuminated using a patterned light field , typically sinusoidal stripes produced by interference of two beams that are split by a diffraction grating .
the resulting image is a product of the illumination pattern and the fluorescence image of the sample . assuming that the dye concentration follows a certain pattern that can be fourier decomposed , one obtains a moir pattern for each component , resulting from the product of the array of illumination lines and the fluorescence signal ( fig .
[ fig_sim]a ) .
the key concept in sim is that the periodicity of a moir pattern is lower than the individual arrays .
let us consider a sample with a periodic array of dyes at distance @xmath30 illuminated by an array of lines spaced by @xmath31 .
if we take the angle of the two line arrays to be zero , the period @xmath32 of the moir pattern is given by @xmath33 . now , consider an objective lens that accepts spatial frequencies @xmath34 .
the highest periodicity moir pattern that is detected by this objective is @xmath35 , so that the decisive criterion becomes @xmath36 furthermore , considering that the highest periodicity illumination that is compatible with the objective is @xmath37 . putting all this information together
, one finds that it is possible to detect fourier components of the sample at @xmath38 . in other words
, one can resolve sample features at a spatial frequency up to twice larger than usual . in resolution ; see text for details . *
( b , c ) * actin cytoskeleton at the edge of a hela cell imaged by conventional microscopy and sim , respectively . *
( d , e ) * insets show that the widths ( fwhm ) of the finest protruding fibers ( small arrows ) are lowered to 110 - 120 nm in * ( c ) * , compared to 280 - 300 nm in * ( b)*. ( b - e : reproduced with permission from @xcite . ) ] by recording a series of images for different orientations and phases of the stripe pattern , one can reconstruct the full image with an improved resolution .
interestingly , the resulting moir pattern has also a three - dimensional structure , which allows the reconstruction program to obtain an enhanced resolution in the axial direction .
figure [ fig_sim]c shows an example of a high resolution image obtained by sim and the comparison to conventional microscopy ( fig .
[ fig_sim]b ) .
there are also approaches that combine structured illumination with other techniques using interference and two opposing objectives to gain high axial resolution ( i@xmath39 m , @xmath40 ) @xcite .
recently , sim has also been combined with light - sheet microscopy ( coined lattice light - sheet microscopy ) @xcite . in an intriguing demonstration ,
3d in - vivo imaging of relatively fast dynamical processes is shown using a bound optical lattice as the light sheet .
it is worth mentioning that sim strategies can not yield an extra factor of two improvement for coherent imaging modalities @xcite .
in 1928 , edward synge proposed to use a thin opaque metal sheet with sub - wavelength holes to illuminate a sample placed at sub - wavelength separation from it @xcite . by scanning the specimen through this point illumination , an image could be recorded with an optical resolution better than the diffraction limit .
technical limitations in the fabrication of nanoscopic apertures , their nano - positioning and sensitive light detection made the experimental realization of a scanning near - field optical microscope ( snom ) only possible in the early 1980s @xcite .
near - field microscopy gets around the diffraction limit in a complete and general fashion .
the essential point is that the limitations imposed by diffraction do not apply to the distances very close to the source , where non - propagating evanescent fields dominate .
these fields contain the high spatial frequency information of the source and sample , but their intensity decays exponentially with a characteristic length of the order of the wavelength of light .
there are two conceptual ways of performing snom . in the first case ,
one follows synge s proposal and illuminates the sample through a subwavelength light source ( see fig . [ fig_nearfield]a ) .
conventionally , the light source is realized by sending light through a tapered fiber that is metalized and has a subwavelength aperture at its end ( see fig . [
fig_nearfield]b ) .
this mode of operation is called aperture snom . here
the size and shape of the aperture dictate the range of spatial frequencies that can be coupled to and scattered by the sample .
the detection can be performed in the far field or back through the aperture although the latter leads to a very unfavorable signal - to - background ratio .
the main limitations of this arrangement are 1 ) small transmission of @xmath41 through the aperture , and 2 ) the fundamental limit of aperture given by the skin depth of metals @xcite . in practice ,
the nanoscopic light source , aperture or scatterer is placed at the end of a sharp tip and various distance regulation mechanisms such as the shear force are used to raster scan it at nanometer separation from the sample .
the most common realization of the aperture snom employs a metal - coated tapered optical fiber with a small aperture of the order of a few nanometers at its apex ( see fig .
[ fig_nearfield]b ) .
however , as mentioned earlier , reaching a resolution below 50 nm proves to be exceedingly hard due to the low tip throughput and the fact that the effective size of the aperture can not be reduced below about 20 nm @xcite . in principle
, this problem can be circumvented by replacing the aperture with a nanoscopic source of light such as a single molecule or a single color center @xcite ( fig .
[ fig_nearfield]c ) .
however , the difficulties of placing the emitter at the very end of the tip and its photostability at room temperature have hampered the wide - spread adoption of this method . in the second snom approach ,
one uses far - field illumination and scans a nanoscopic scatterer or antenna very close to the sample ( cf .
[ fig_nearfield]d ) .
the near field of this nanostructure again contains high spatial frequencies , which can be scattered by the sample . in general , this so - called apertureless snom mode @xcite is much more challenging .
the main difficulty is that the far - field illumination creates a large background from an area ( minimum of a diffraction - limited area ) that is much larger than the near - field domain of the antenna .
furthermore , this background light easily results in interferometric artifacts when the height of the antenna or sample are changed @xcite .
the most common platform for performing apertureless snom has been the solid tip , which might be metalized @xcite ( fig . [
fig_nearfield]e ) .
although such experiments in the infrared domain have become well established @xcite , reproducible fabrication and efficient characterization of the suitable tips have made apertureless snom in the visible regime an uphill battle . as a result , many scientists turned to using more well - defined metallic nanoparticles placed at the end of a dielectric tip@xcite ( fig .
[ fig_nearfield]f ) .
this approach has been particularly successful in the context of antenna - based snom and in producing quantitative data on the near - field enhancement of fluorescence from single molecules @xcite .
an alternative way of imaging in the near field has emerged in the context of metamaterials .
these artificial materials are structures with sub - wavelength - sized unit cells and can be engineered to exhibit intriguing properties such as a negative index of refraction @xcite . in the year 2000 , a so - called perfect lens was proposed by john pendry using a slab of negative index material @xcite . as illustrated in fig .
[ fig_nearfield2]a , the idea with such a lens is that an object that is embedded in a material with refractive index @xmath22 can be perfectly imaged by a slab of a material with refractive index @xmath42 , assuming that it is perfectly impedance matched and completely lossless .
the exponential decay of the evanescent wave intensity inside the @xmath22 material is completely reversed by an exponential amplification inside the material with @xmath42 refractive index . thus both propagating and evanescent fields are imaged by this lens yielding a perfect image .
such a perfect lens has not yet been experimentally realized because of the extremely delicate constraints on the properties of the negative index material .
there are , however , experimental demonstrations of a superlens in the optical regime where sub - diffraction imaging was shown using metamaterial structures @xcite .
there have also been efforts for the realization of a hyperlens ( see fig .
[ fig_nearfield2]b ) to project the near field into the far field using a cylindrical @xcite or a spherical hyperlens @xcite .
fabrication issues , material properties and the requirement that the object of interest must be placed in the near field of the hyperlens , make the practical usage of these interesting imaging techniques very limited although they might possibly find use in nanofabrication and optical data storage .
fluorescence microscopy witnessed several important developments in the 1980s .
this included commercialization of scanning confocal microscopy and the invention of two - photon absorption microscopy .
a young doctoral student in heidelberg , stefan hell , took upon himself not to accept the diffraction limit in its usual formulation . shortly after finishing his doctoral work , hell proposed to exploit stimulated emission to deplete ( sted ) the fluorescence of molecules in the outer part of the illumination and thereby reduce the size of the effective fluorescence spot @xcite .
the first experimental realization of this idea was reported just a few years later @xcite .
hell was awarded the nobel prize in chemistry in 2014 for his achievements in this area . upon excitation
, a fluorescent molecule is usually brought from its singlet ground state s@xmath43 to a higher vibrational state of the singlet electronic state s@xmath44 ( see fig .
[ fig_sted]a ) , which then relaxes on a picosecond time scale to the lowest vibrational level .
if the quantum efficiency of the molecule is high , a photon is emitted within a few nanoseconds .
hell had the idea to suppress this fluorescence in the outer part of the excitation beam by stimulating the emission much faster than the nanosecond spontaneous emission in that region . to do this , he used a doughnut - shaped laser beam profile ( see fig .
[ fig_sted]b ) that is overlapped on the excitation focal spot .
the stimulated emission takes place at a wavelength that is red - shifted with respect to the main part of the fluorescence line , allowing one to spectrally separate the stimulated emission from various fluorescence components .
it is important to note that the depletion doughnut beam itself is also diffraction limited , but the effective size of its hole in the middle can be adjusted by the beam intensity .
in other words , the higher the intensity of the depletion beam , the farther one goes into the saturation of the fluorophores . as a result of this nonlinear behavior
, the fluorescence psf can be sculpted .
it follows that the resulting sub - diffraction resolution can be described by a modified form of ernst abbe s equation after considering the degree of saturation , @xmath45 here , @xmath46 is the peak intensity of the depletion laser and @xmath47 denotes the saturation intensity of the fluorophore .
the resolution becomes sub - diffraction limited when the ratio between @xmath46 and @xmath47 becomes larger than one .
the best resolution that has been demonstrated with sted is about 20 nm in the case of standard fluorescence labeling assays with organic fluorophores @xcite and about 50 nm using genetically expressed fluorescent proteins in live cells @xcite . in the case of a very robust fluorophore such as a nitrogen - vacancy color center in diamond , a resolution down to 2 - 3 nm
was successfully demonstrated @xcite .
sub - diffraction imaging by using a doughnut - shaped excitation pattern for fluorescence suppression can be generalized to other photophysical mechanisms besides stimulated emission .
for example , one can exploit the saturated depletion of fluorophores , which can be reversibly switched between a bright state and a dark state .
this dark state can have a variety of origins such as the ground state of a fluorophore in sted , its triplet state in ground - state depletion ( gsd ) microscopy @xcite , or a nonfluorescent isomer of an over - expressed fluorescent protein @xcite .
this general principle was coined `` resolft '' as an abbreviation for reversible saturable optically linear fluorescence transitions , which is also a pun on the way germans might pronounce `` resolved '' @xcite .
the resolft concept can also be applied to sim , also known as saturated sim ( ssim ) @xcite . in this regime ,
a sinusoidal illumination pattern becomes effectively more and more rectangular as the excitation intensity increases .
this leads to higher order fourier terms in the illumination pattern periodicity @xmath31 . using similar back - of - the - envelope considerations as for standard sim , @xmath48
, one can show that spatial frequencies @xmath49 become detectable .
indeed , using ssim , lateral spatial resolutions on the order of 50 nm have been demonstrated @xcite .
the general class of resolft methods has offered a very clever strategy for circumventing the diffraction limit in fluorescence microscopy . nevertheless , these methods are accompanied with challenges , which will call for more innovations in the years to come .
one of the issues is the fact that so far resolft has been a raster scan technique with a certain temporal resolution , which might limit live cell imaging applications or the study of dynamic processes .
this issue can be resolved to some extent by massive parallelization @xcite , e.g. via a square grid similar to structured illumination microscopy .
a second challenge concerns the photophysics of the fluorophores .
excitation to higher electronic states can efficiently compete with the stimulated emission process , opening pathways for photobleaching @xcite .
the company abberior has addressed this problem by developing a range of suitable dye molecules and other fluorophores covering the greater part of the visible spectrum @xcite
. a further area of development will be multicolor applications as it is common in biological fluorescence microscopy .
since already two lasers are necessary for resolft methods , the implementation of two or more color channels is somewhat more complex than in standard fluorescence microscopy .
nevertheless , sted with two different fluorescence labels and two pairs of excitation and sted lasers has been demonstrated @xcite .
there have also been efforts for sharing the excitation and sted lasers @xcite , or separating a third fluorophore that lies in the same spectral region by its fluorescence lifetime @xcite .
imaging deep in the sample also poses difficulties for resolft microscopy . as the attainable resolution critically depends on the quality of the intensity minimum and phase fronts in the center of the depletion beam , slight changes of the depletion beam profile by scattering deteriorates the imaging performance .
this problem can be , in principle , addressed by recent developments using adaptive optics @xcite . in parallel to the developments of snom in the 1980s and 1990s , scientists worked hard to detect matter at the level of single ions and single molecules .
these efforts were originally more motivated by fundamental issues in spectroscopy and to a good part by the community that had invented methods such as spectral hole burning for obtaining high - resolution spectra in the condensed phase .
although already in the 1970s and 80s there had been indications of reaching single - molecule sensitivity in fluorescence detection @xcite , the first reliable and robust proof came from the work of w. e. moerner , who showed in an impressive experiment that a single pentacene molecule embedded in an organic crystal could be detected in absorption at liquid helium temperature @xcite . soon after that m. orrit demonstrated a much better signal - to - noise ratio by recording the fluorescence signal in the same arrangement @xcite .
the ease of this measurement kick - started the field of single molecule fluorescence detection .
however , it was not until 1993 that e. betzig provided the first images of single molecules at room temperature @xcite .
betzig used a fiber - based aperture snom to excite conventional dye molecules on a surface . at this point
, there was a strong belief that far - field excitation would not be favorable because it would cause a large background .
interestingly , shortly after that r. zare s lab demonstrated scanning confocal images of single molecules @xcite .
this achievement was the final step towards a widespread use of single molecule fluorescence microscopy .
a particularly interesting feature of fluorescence that was revealed by single molecule detection is photoblinking , i.e. the reversible transition between bright and dark states of a fluorescent molecule .
different physical mechanisms may cause fluorescence intermittencies depending on the type of fluorophore as well as its surroundings @xcite .
for example , an excited molecule can undergo a transition to a metastable triplet state with a much longer lifetime than the singlet excited state ( see fig .
[ fig_triplet ] ) . during this time , the molecule is off because it can not be excited .
the evidence for triplet state blinking is an off - time distribution that follows an exponential law .
however , in some cases the off - time statistics reveal a power law similar to the blinking observed in semiconductor nanocrystals @xcite .
a proposed mechanism for such a fluorescence intermittency is the formation of a radical dark state @xcite .
there have been several studies on the topic of blinking , but there is only a limited amount of room - temperature and low - temperature data available and many questions remain open @xcite .
fluorophores also undergo photobleaching , i.e. an irreversible transition to a non - fluorescent product . at room temperature dye molecules
typically photobleach within several tens of seconds or a few minutes if sophisticated antifading reagents are used in the buffer .
however , the survival times at cryogenic temperatures can go beyond an hour @xcite or even days in the case of a crystalline matrix .
in the special case of terrylene in p - terphenyl a comparable photostability has been achieved even at room temperature @xcite .
unfortunately , the combination of aromatic molecules and crystalline host matrices is not compatible with the labeling strategies in the life sciences .
the idea behind localization microscopy is to find the position of each fluorophore by determining the center of its diffraction - limited psf with a higher precision than its width .
this is accomplished by fitting the distribution of the pixel counts on the camera with a model function that describes this distribution ( see fig . [ fig_sml ] ) .
the principle was already conceived by werner heisenberg in the 1920s @xcite and experimentally demonstrated in the 1980s in the context of localizing a single nanoscale object with nanometer precision @xcite . in this scheme , single emitters can be localized with arbitrarily high precision only dependent on the available snr .
the localization precision is mainly determined by the number of photons that reach the detector , the size of the psf , the level of background noise and the pixel size @xcite .
the background noise is in turn affected by the luminescence of the cover slip or other materials on the sample as well as the dark counts and readout noise of the camera @xcite . the attainable localization precision ( @xmath50 )
can be written as @xmath51 using a maximum likelihood estimation procedure with a 2d gaussian function .
this predicts a localization error close to the information limit @xcite .
here , @xmath52 denotes the detected number of photons , @xmath53 stands for the half - width of the psf given by the standard deviation of a gaussian profile , @xmath54 is the level of background noise and @xmath55 denotes the pixel size .
the limiting factor is typically the finite value of @xmath52 caused by irreversible photobleaching of the fluorophore .
the photon budget of commonly used photoactivatable fluorescent proteins lies in the range of a few hundred detected photons @xcite , which typically leads to a localization precision on the order of 20 nm . to improve on this limitation
, several efforts have optimized the choices of fluorophores and the buffer conditions @xcite , engineered the dye molecule itself @xcite , or carefully controlled its environment @xcite .
the best localization precision for single molecules that has been reported is just under 0.3 nanometers @xcite .
a particularly powerful tool based on the concept of localization is single particle tracking .
localizing a fluorescent marker or non - fluorescent nano - object of interest as a function of time allows one to study dynamical processes such as diffusion in lipid membranes @xcite .
single particle tracking has been performed with various imaging modalities including fluorescence @xcite , scattering @xcite and absorption @xcite .
however , high temporal and spatial precisions call for a trade - off because smaller integration times lead to a lower signal and lower snr .
interferometric scattering microscopy ( iscat ) can provide an ideal solution @xcite , offering up to mhz frame rates in combination with nanometer localization precision in the case of small scatterers like gold nanoparticles .
interestingly , localization techniques and particle tracking have also found applications in a wide range of studies such as coherent quantum control @xcite or cold atoms @xcite .
for example , identification of atoms down to the single lattice site of an optical lattice has provided an avenue to manipulating single qubits and studying many - body effects like the quantum phase transition from a superfluid to a mott insulator .
given that a single molecule can be localized to an arbitrarily high precision , one could also resolve two nearby molecules if only one could address them individually .
this was formulated by e. betzig in 1995 as a general concept @xcite , but it was already demonstrated experimentally in 1994 by the group of urs wild in cryogenic studies @xcite . in the latter ,
the inhomogeneous distribution of the molecular resonance lines allows one to address each molecule separately by tuning the frequency of a narrow - band excitation laser .
we have recently demonstrated that the same principle of spectral selection can also be used to address single ions in a solid @xcite . by combining cryogenic high - resolution spectroscopy and local electric field gradients ,
it was also shown in our laboratory that two individual molecules could be three - dimensionally resolved with nanometer resolution @xcite . to our knowledge
, those results still establish the highest three - dimensional optical resolution .
several groups have tried different strategies for distinguishing neighboring fluorophores .
one example was to use semiconductor nanocrystals with different emission spectra @xcite or stepwise bleaching of single molecules @xcite .
however , extension of these methods to very large number of fluorophores was not practical .
the decisive breakthrough came in 2006 , when three groups reported very similar strategies based on stochastic photoactivation processes that switched the fluorophores between a dark state and a fluorescent state ( see also fig . [ fig_storm]b ) .
eric betzig and colleagues called their method photoactivated localization microscopy ( palm ) @xcite , sam hess and his team coined the term fluorescence photoactivation localization microscopy ( fpalm ) @xcite , and x. zhuang and her group used the term stochastic image reconstruction microscopy ( storm ) @xcite . in each case , the fluorophores are placed on the target structure by different labeling techniques . while commonly used antibody - based assays have a label - to - target distance of about 20 nm , using nanobodies @xcite , aptamers @xcite or fluorescent proteins @xcite can reduce that distance to a few nanometers .
m , main image ; 100 nm , insets .
( c , d : reproduced with permission from @xcite . ) ] figure [ fig_storm ] illustrates the data acquisition procedure for super - resolution imaging based on single molecule localization . by shining light on the sample with a blue - shifted activation laser beam
, one can stochastically switch on a sparse subset of fluorophores .
next , one turns on the excitation laser and collects the fluorescence from the few activated fluorescent molecules until they become deactivated . by adjusting the intensity of the activation beam
, one can control the average number of activated fluorescent labels to ensure that psfs from individual fluorophores do not overlap .
one then performs a localization analysis for each recorded psf to determine the positions of all molecules .
the process of activation , recording and localization is then repeated for many other random subsets of fluorophores until one is satisfied with the number of labels for reconstructing a super - resolution image .
there are also variations of this acquisition procedure using , for instance , asynchronous activation and deactivation of fluorophores @xcite or assays where diffusing fluorophores get activated upon binding to the target structure @xcite . in the standard super - resolution imaging modalities such as palm and storm , usually photochemistry is employed in order to exert some degree of control on the photoswitching kinetics .
this control is necessary since the achievable resolution also depends on the ratio of the on- and off - switching rates of the used fluorophores @xcite .
examples of such photochemistry is the chromophore cis - trans isomerization or protonation change in the case of fluorescent proteins @xcite , and the interplay of reduction and oxidation using enzymatic oxygen - scavenging systems and photochromic blinking for organic dye molecules @xcite .
the state of the art in resolution for localization microscopy is about 10 nm @xcite limited by the total number of photons emitted before photobleaching .
however , even sub - nanometer localization precision has been reported in cases where photobleaching could be delayed by using oxygen scavengers @xcite or cryogenic conditions@xcite .
the latter measurements offer the additional advantage of a more rigid sample fixation than chemical fixatives .
indeed , first studies exploring palm imaging at low temperature have also recently surfaced @xcite .
the full arsenal of methodologies developed for cryogenic electron microscopy may be applied to prepare samples for cryogenic super - resolution microscopy and even dynamic processes could be studied either by stopping processes at different times or by employing methods like local heating with an infrared laser @xcite .
a quantitative assessment of the molecules positions critically depends on both the precision and accuracy of the employed method .
precision in localization microscopy is determined by the standard deviation of the estimated position of an emitter assuming repeated measurements , whereas the accuracy quantifies how close the estimated position lies to the true position . in other words , even if the measurement precision is high , absolute distance information might be compromised by technical sources of bias like pixel response non - uniformity of the camera or sample drift @xcite .
an important systematic source of error concerns the dipolar emission characteristics of single molecules .
it is known that the image of an arbitrarily oriented dipolar light source deviates from a simple isotropic psf ( see fig .
[ fig_dipoleorientation ] ) . as a result ,
localization of the individual fluorophores requires a fitting procedure that takes this effect into account @xcite . in the presence of nearby interfaces
, this can become a nontrivial task @xcite .
however , the psf asymmetry is much less pronounced if a microscope objective with low numerical aperture is used @xcite . of course , fitting the data with full theoretical treatment of the psf or good approximations provides accurate values for the position and orientation of the fluorophore even in the case of high numerical aperture @xcite . the most severe limit in localization microscopy is the difficulty of high - density labeling .
here it is to be remembered that the image in this method is constructed by joining the centers of the individual fluorophores ( see fig .
[ fig_storm ] ) .
this means that a resolution of 5 nm in deciphering the details of a figure would need at least two fluorophores that are spaced by about 2.5 nm according to the sampling theorem @xcite .
the first problem with this requirement is the difficulty of labeling at such high densities .
second , once one manages to place the fluorophores at the right place , one faces the problem that such closely - spaced fluorophores undergo resonance energy transfer ( homo - fret ) @xcite . as a result , the emission can not be attributed to one or the other fluorophore , and the basic concept of localization microscopy breaks down .
another issue to be considered is that of accidental overlapping psfs from neighboring fluorophores . in high - precision co - localization microscopy
this situation can usually be avoided by rejecting the affected psfs .
however , it would be more advantageous to localize these psfs as well .
it turns out that the fluorophore density can be increased to achieve faster acquisition times by using appropriate algorithms @xcite .
and the distance between the molecule and the interface is @xmath56 . *
( b ) * examples for psfs for two different numerical apertures and four inclination angles at @xmath56 = 2 nm . ]
the lateral resolution in the above - mentioned super - resolution methods is currently of the order of 10 - 50 nm , but these techniques are not fundamentally limited by any particular physical phenomenon .
the resolution is rather hampered by practical issues , which can be addressed to various degrees in different applications .
thus , only emphasizing a record resolution outside a specific context is not a meaningful exercise .
an important development concerns super - resolution in three dimensions .
one possibility to localize a fluorophore along the optical axis is via astigmatism @xcite . by inserting a cylindrical lens in the detection path ,
the psf becomes elliptical , from the degree of its ellipticity and orientation one can deduce the additional axial position of the fluorophore .
lateral localization precision of about 25 nm with an axial localization precision of about 50 nm was already reported in 2008 @xcite .
recently , an isotropic localization precision of about 15 nm in all three spatial dimensions was reported using storm in combination with an airy - beam psf @xcite . an alternative way to obtain 3d
super - resolution is multi - focal plane imaging @xcite . here
, different focal planes are imaged on various regions of the camera by splitting the fluorescence light and introducing different path lengths .
the height can then be deduced from the degree of defocusing .
another approach uses an engineered psf that encodes the axial position of the emitter in the rotation angle of two lobes , a double - helix psf @xcite .
a crucial requirement for practical biological microscopy is the ability to image many entities simultaneously .
the most convenient and common approach is to label different parts of the specimen with fluorophores of distinct absorption or emission spectra . considering that the new super - resolution methods , including resolft , palm , storm , etc .
, all rely on the photophysical properties of the fluorophores , it is not a trivial matter to marry them with multicolor imaging . first , fluorescence probes with the desired switching properties must be available with the correct excitation , emission and activation wavelengths .
interestingly , scientists began to develop multicolor super - resolution solutions shortly after the introduction of localization microscopy .
indeed , two - color @xcite and even three - color super - resolution imaging has been demonstrated @xcite .
a second challenge concerns the crosstalk that is caused by the spectral overlap of the emission bands of different fluorophores , which are typically several tens of nanometers broad at room temperature .
a possible solution would be to perform super - resolution microscopy at low temperatures because even though dye molecules suited for labeling in life science do not show lifetime - limited linewidths at cryogenic temperatures , their spectra can become narrower by orders of magnitude . as super - resolution optical microscopy becomes a workhorse
, it becomes more and more important that the methods can also handle live cell imaging . here
, imaging speed and phototoxicity pose important problems . on the one hand
, fast imaging often requires a large excitation dose to be able to collect lots of photons in a short time window .
high light dose , however , causes the production of free radicals through the photo - induced reaction of the fluorophore with molecular oxygen @xcite .
furthermore , excess light also brings about fast photobleaching and short observation times .
an interesting approach to minimize these phototoxic effects is light - sheet microscopy @xcite . here ,
the wide - field detection is disentangled from the illumination , which consist of a thin sheet of light perpendicular to the focal plane . by performing tomographic recordings at different sample orientations , one
can then obtain impressive three - dimensional images of whole organs in small animals such as zebrafish @xcite .
of course , diffraction limits the thickness of the light sheet to dimensions well above a wavelength , especially if the illumination area is to be large . by employing slowly diffracting beams such as bessel beams
, one can minimize the problem of beam divergence @xcite .
light - sheet microscopy is very popular in developmental biology , where super - resolution is less important than large - scale information about the whole system over a longer time .
an application example of the technique is the four - dimensional imaging of embryos at single - cell resolution @xcite .
if we now relax the requirements for routine biological microscopy , we find that several experiments have already extended super - resolution microscopy to the nanometer and even sub - nanometer level .
the first demonstration of nanometer resolution in all three spatial dimensions used low - temperature fluorescence excitation spectroscopy in combination with recording the position - dependent stark shift of the molecular transition in an electric field gradient @xcite . in another experiment
, a distance of about 7 nm was measured between two different dye molecules with an accuracy of 0.8 nm using a feedback loop for the registration of two color channels and oxygen - reducing agents@xcite .
the most recent achievement concerns angstrom accuracy in cryogenic colocalization . here , two identical fluorophores attached to a double - stranded dna at well - defined separations as small as 3 nm were resolved with a distance accuracy better than 1 nm @xcite .
figure [ fig_coloc]a illustrates how this can be used to determine the positions of the dye molecules and their separation .
the intensity trace retrieved from the image stack of a dna molecule with the attached fluorescent molecules shows three levels .
the upper level corresponds to the state where both fluorophores are on , the middle one to the case that only one is fluorescent , and in the lowest level both fluorophores are in a dark state .
after localizing the single fluorophores in the frames where there is only one present , one can subtract the image frames from those where both fluorophores were on . in this fashion
, one can also localize the second fluorophore .
an alternative route is to find the center of mass of each fluorophore and then compute the distance . using the latter approach ,
both sub - nanometer localization precision and accuracy for the distance measurement have been demonstrated ( see fig . [ fig_coloc]b ) .
the holy grail of super - resolution microscopy is to break free from the shackles of fluorescent labels .
the use of fluorescence markers introduces a variety of difficulties , starting with the labeling process itself and ending with the inevitable photobleaching of the labels . in the past years
, there have been several proof - of - principle demonstrations to obtain fluorescence - free super - resolution images .
one possibility is a technique called optical diffraction tomography ( odt ) @xcite . here
, the sample is illuminated using every possible angle of incidence allowed by the numerical aperture of the microscope objective .
then the intensity , phase and polarization state of the scattered far field are recorded for different angles , and the distribution of the permittivity of the object of interest is reconstructed numerically .
recently , an optical resolution of about one - fourth of the wavelength was experimentally demonstrated @xcite .
label - free super - resolution has also been demonstrated using surface - enhanced raman scattering ( sers ) @xcite , where one performs a stochastic reconstruction analysis on the temporal intensity fluctuations of the sers signal .
another intriguing method employs ground - state depletion of the charge carriers with a doughnut shaped beam resulting in the transient saturation of the electronic transition by using a pump - probe scheme @xcite .
there have also been several approaches to achieve super - resolution imaging using the photon statistics of the emitters @xcite .
as discussed earlier , it is also possible to detect unlabeled biomolecules such as proteins via iscat detection of their rayleigh scattering @xcite . in this method
the image of a single protein can be localized in the same manner as in fig .
[ fig_sml ] . here
too , one needs to turn the proteins on and off individually if one wants to resolve them beyond the diffraction limit . in dynamic experiments ,
the arrival time of each protein can serve as a time tag @xcite . the localization precision and
therefore the attainable resolution is determined by the signal - to - noise ratio , which in turn depends on the size of the biomolecule in iscat .
this size - dependent signal also provides a certain level of specificity that in fluorescence modalities can only be achieved by employing different fluorophores .
another report has used the saturation of scattering in plasmonic nanoparticles .
although the saturation effect in the absorption of small plasmonic nanoparticles has been studied for many years @xcite , saturation of scattering has only been reported recently @xcite .
this saturation stems from a depletion of the plasmon resonance .
similar to the case of super - resolution imaging in ssim , the saturation effect in scattering allows one to record images with a resolution beyond the diffraction - limit . by recording images at different light intensities ,
super - resolved images were obtained and a resolution of @xmath57 was demonstrated @xcite .
finally , combination of optical microscopy with other imaging modes such as scanning probe techniques or electron microscopy can offer very useful additional information about the sample .
some of the recent examples of correlative microscopy are the combination of optical super - resolution microscopy with electron microscopy @xcite and with atomic force microscopy ( afm ) @xcite .
the quest for inventing new imaging mechanisms and pushing the spatial and temporal resolutions is a fundamental challenge for physicists and of great practical importance in science and technology .
abbe s formulation of the diffraction limit at the end of the nineteenth century put a harsh spell on optical microscopy , which lasted for about one hundred years .
the advent of scanning near - field microscopy broke this spell , and once the dogma of a fundamental limitation of resolution was eliminated , scientists reconsidered many scenarios and explored fascinating techniques , which we have discussed in this review article .
it is now fully accepted that resolving two small objects at very close distances could be , in principle , achieved with an arbitrary resolution and accuracy . the key concept in breaking
the diffraction barrier has been to exploit more information from the system , e.g. taking advantage of the spectroscopic energy levels or transitions of fluorophores . in this spirit
, scientists continue to develop optical imaging methods by devising clever schemes that rely on nonlinear phenomena , quantum optics or ultrafast laser spectroscopy .
the many elegant ideas do not , however , all have practical implications .
in particular , one has to consider many restrictions in biological imaging .
for example , the amount of laser power that one can shine onto a live cell before it is damaged is orders of magnitude lower than what a diamond sample can take .
moreover , there are important issues concerning labeling techniques and the influence of the label on the functionality of its environment .
one subtle point regards the production of free radicals in a photochemical reaction of the excited fluorophore with the surrounding oxygen molecules . to minimize the effect of phototoxicity , it is helpful to acquire images as efficiently as possible
of course , this is also highly desirable because one gets access to more of the dynamics of the biological and biochemical processes .
the nobel prize in chemistry in 2014 has honored the contributions of optical scientists in the area of super - resolution fluorescence microscopy and single - molecule detection .
this is the fourth prize after dark - field microscopy ( zsigmondy 1925 ) , phase contrast microscopy ( zernike 1953 ) , and the green fluorescent protein ( chalfie , shimomura and tsien 2008 ) , which is dedicated to optical microscopy , spread over nearly one hundred years .
the recent achievements are a testimony to the livelihood of light microscopy as a research field in fundamental science .
they show a new trend against the older belief that the physics of imaging is fully understood and that its development belongs to engineering departments .
we are convinced that the combination of concepts from laser spectroscopy , quantum optics , photophysics , photochemistry , nanotechnology , and biophysics will introduce many new avenues for optical imaging
. currently resolution at around 10 - 50 nm is routinely reported in various configurations , but we have shown that this limit can be pushed by another one hundred times to the sub - nanometer level . more importantly various imaging contrasts , e.g. label - free techniques , promise to open the door to whole new classes of information .
finally , temporal resolution will be an area of innovation and growth .
the dynamics of interest in biomedical processes range from femtosecond for electronic and vibrational degrees of freedom to days and years for growth and disease progression .
this astronomical span of the time scales will certainly require totally different techniques .
progress in all these cases will ultimately confront the _ signal - to - noise barrier_. development of methods for efficient collection of photons and their combinations with lab - on - chip solutions , advances in photochemistry and photophysics of new labels as well as better detector and laser technologies will all contribute to pushing this barrier .
experimental physicists and in particular optical scientists are well positioned to lead the ongoing revolution of optical microscopy if they manage to achieve a high degree of cross fertilization between biology , chemistry , medicine and physics .
we thank anna lippert , richard w. taylor , simon vassant and luxi wei for discussions and critical comments on the manuscript . we acknowledge support from the alexander von humboldt foundation and the max planck society . | optical microscopy is one of the oldest scientific instruments that is still used in forefront research .
ernst abbe s nineteenth century formulation of the resolution limit in microscopy let generations of scientists believe that optical studies of individual molecules and resolving sub - wavelength structures were not feasible . the nobel prize in 2014 for super - resolution fluorescence microscopy marks a clear recognition that the old beliefs have to be revisited . in this article , we present a critical overview of various recent developments in optical microscopy .
in addition to the popular super - resolution fluorescence methods , we discuss the prospects of various other techniques and imaging contrasts and consider some of the fundamental and practical challenges that lie ahead .
[ 1995/12/01 ] |
the strong electron correlation in high - temperature superconductors ( htsc ) is able to drive an electronic nematic phase @xcite , which preserves translational symmetry but breaks rotational symmetry . in the past decade , there have been a number of experimental signatures pointing to the presence of nematic ordering transition in some htscs @xcite .
on the basis of these experiments , a zero - temperature nematic quantum critical point ( qcp ) is supposed to exist at certain doping concentration @xmath3 in the superconducting ( sc ) dome @xcite . generally , the nematic order has two impacts on the sc state .
first , it competes with the sc order @xcite .
second , the nematic order parameter couples to the gapless nodal quasiparticles ( qps ) , which are believed to be the most important fermionic excitations in unconventional superconductors with @xmath4 energy gap .
the latter coupling is singular at the nematic qcp @xmath3 , and has stimulated considerable theoretical efforts @xcite .
a recent renormalization group analysis @xcite showed that it leads to a novel fixed point at which the ratio between gap velocity @xmath5 and fermi velocity @xmath6 of nodal qps flows to zero , @xmath7 .
although a zero - temperature nematic qcp is expected to exist somewhere in the sc dome @xcite , shown schematically in fig .
( [ fig:1 ] ) , its precise position , and even its very existence , has not been unambiguously confirmed by experiments so far .
it is therefore always interesting to seek evidence which can help convincingly confirm or disconfirm the existence of such point . in this paper
, we study the superfluid density @xmath1 and the sc temperature @xmath2 at the supposed nematic qcp @xmath3 . if @xmath1 and @xmath2 exhibit sharply distinct behaviors at @xmath3 , then the nematic qcp may be detected by measuring these quantities .
htscs are known to be doped mott insulators , so their superfluid density is much smaller than that of conventional metal superconductors . at @xmath0 , the superfluid density in underdoping region depends @xcite linearly on doping @xmath8 as @xmath9 , where @xmath10 is the lattice spacing . at finite @xmath11 ,
certain amount of nodal qps are thermally excited out of the sc condensate .
lee and wen argued that these normal nodal qps can efficiently deplete the superfluid density @xcite .
formally , the superfluid density contains two terms , @xmath12 , where @xmath13 is the normal qps density .
setting @xmath14 allows for an estimate of the critical temperature @xmath2 . employing a phenomenological approach , lee and
wen @xcite obtained @xmath15 , reproducing the uemura plot @xcite .
once a nematic ordering transition occurs at @xmath3 , the superfluid density and @xmath2 will be substantially changed .
as @xmath16 due to the critical nematic fluctuations , it seems that @xmath17 , i.e. , superconductivity would be completely suppressed at @xmath3 .
this argument is certainly oversimplified since the above expression of @xmath2 is obtained in the non - interacting limit .
however , this qualitative analysis does indicate the importance of the critical nematic fluctuations , and indeed motivates us to perform a quantitative computation of the renormalized @xmath18 and @xmath2 after taking into account the nematic fluctuations .
the nematic order affects @xmath18 in two ways . on the one hand , since the nematic order competes with the sc order , it reduces @xmath19 .
this reduction can be examined by studying the competitive interaction between nematic and sc order parameters . on the other
, the critical nematic fluctuations can excite more nodal qps out of the condensate , compared with the case without nematic order . as a consequence ,
@xmath13 is enhanced and the superfluid density is further suppressed .
we shall access this effect by generalizing the phenomenological approach proposed in ref .
the velocity anisotropy plays an essential role in these considerations .
after explicit calculations , we find that superfluid density @xmath1 and @xmath2 are both significantly reduced due to critical nematic fluctuations , indicating a strong suppression of superconductivity at nematic qcp @xmath3 ( see fig .
( [ fig:1 ] ) ) . at nematic qcp @xmath3
is represented by the dashed line.,width=316 ] the rest of the paper is organized as follows . in sec.ii
, we address the competitive interaction between sc and nematic order parameters and calculate zero-@xmath11 superfluid density . in sec.iii , we calculate the density of normal qps after taking into account fermion velocity renormalization due to critical nematic fluctuation . based on these calculations ,
we predict a strong suppression of superconductivity at nematic qcp . in sec.iv
, we present a brief summary of our results , and also discuss the possible experimental determination of the nematic qcp .
we first consider the renormalized zero-@xmath11 superfluid density at nematic qcp . based on phenomenological grounds , we write down a free energy of two competing orders , @xmath20 where @xmath21 and @xmath22 are the sc and nematic order parameters , respectively .
the strength of the competitive interaction between @xmath21 and @xmath22 is represented by a positive constant @xmath23 .
such type of effective model has been adopted to describe competing orders in various superconductors @xcite . in the absence of nematic order ,
the mean value of @xmath21 is @xmath24 . to be consistent with experiments ,
the parameters @xmath25 and @xmath26 must be properly chosen such that @xmath27 . in the presence of nematic order
, @xmath28 will be renormalized by the @xmath29 term .
the quantum fluctuation of nematic order @xmath22 is very strong and actually singular at nematic qcp @xmath3 , so @xmath22 should be regarded as a quantum - mechanical field .
however , we can consider @xmath21 as a classical field and ignore its quantum fluctuations , provided @xmath3 is not close to the sc qcp @xmath30 .
the free energy of nematic order , @xmath31 , should be specified now .
analogous to the free energy of sc order @xmath21 , @xmath31 contains a quadratic term @xmath32 and a quartic term @xmath33 .
however , the additional coupling between nematic order and nodal qps introduces an extra term .
the action describing this coupling is given by @xcite @xmath34 , \\
s_{\psi\phi } & = & \int d^2xd\tau [ \lambda_0 \phi(\psi^{\dagger}_{1i}\tau^{x}\psi_{1i } + \psi^{\dagger}_{2i}\tau^{x}\psi_{2i})],\end{aligned}\ ] ] where @xmath35 are pauli matrices and the flavor index @xmath36 sums up @xmath37 to @xmath38 .
@xmath39 represents nodal qps excited from @xmath40 and @xmath41 points , and @xmath42 the other two .
the effective action of @xmath22 has the form @xcite @xmath43,\label{eq_effective_phi}\end{aligned}\ ] ] which is obtained by integrating out @xmath38-flavor nodal qps . here , @xmath44 is the tuning parameter for the nematic ordering transition , with @xmath45 defining the nematic qcp @xmath3 .
the polarization function @xmath46 comes from nodal qps , and , to the leading order of @xmath47-expansion , is defined as @xmath48 , \nonumber \\\end{aligned}\ ] ] where @xmath49 is the free propagator for nodal qps @xmath50 ( free propagator for nodal qps @xmath51 can be similarly written down ) .
the polarization function @xmath52 has already been calculated previously @xcite , and is known to have the form @xcite @xmath53 . \nonumber
\\\end{aligned}\ ] ] note there is no direct interaction between @xmath21 and nodal qps @xmath54 .
indeed , @xmath54 are excited on top of a sc order @xmath21 .
moreover , they have sharp peak and long lifetime in the sc dome in the absence of competing orders @xcite , so their coupling to @xmath21 must be quite weak .
starting from eq .
( [ eq_free - energy ] ) and eq .
( [ eq_effective_phi ] ) , we can compute the correction to zero-@xmath11 superfluid density due to the competition between sc and nematic orders . to this end , we need to minimize the effective potential ofsc order , @xmath55 $ ] .
the bare potential for @xmath21 is simply @xmath56 = - \alpha \psi^2 + \frac{\beta}{2}\psi^4 $ ] .
it receives an additional term @xmath57 $ ] due to the sc - nematic competition .
this additional term will be calculated using the methods presented in refs .
the corresponding partition function is @xmath58 = \int \mathcal{d}\phi(\mathbf{r},\tau ) \exp\left(-\frac{\mathcal{f_{\psi}}}{t } - s_{\phi } - s_{\psi\phi}\right)\end{aligned}\ ] ] where @xmath59 .
the saddle - point equation for @xmath21 reads @xmath60}{\delta\psi(\mathbf{r } ) } = 0,\end{aligned}\ ] ] which gives rise to @xmath61\right]\psi(\mathbf{r } ) = 0.\end{aligned}\ ] ] here , @xmath62 $ ] is the expectation value of @xmath32 . at the one - loop level , it has the form @xmath63 \equiv \langle \phi^2 \rangle = \int \frac{d^3 q}{(2\pi)^3 } \frac{1}{q^2 + \pi(q ) + \gamma\psi^2},\end{aligned}\ ] ] where @xmath46 and @xmath64 represent the contributions due to nodal qps and sc order , respectively .
although the @xmath65 term appearing in the denominator of @xmath62 $ ] is much smaller than @xmath46 at low energy , it can not be simply neglected since @xmath62 $ ] would be divergent without such term .
it is now straightforward to get @xmath66}{\delta \psi } = 2\gamma\psi f[\psi],\end{aligned}\ ] ] which then leads to a renormalized potential @xmath67 = - \alpha \psi^2 + \frac{\beta}{2}\psi^4 + v_1[\psi].\end{aligned}\ ] ] to calculate the renormalized zero-@xmath11 superfluid density , @xmath68 , one needs to minimize the effective potential @xmath55 $ ] by taking @xmath69}{\delta \psi } = 0.\end{aligned}\ ] ] the renormalized @xmath70 and therefore @xmath68 can be obtained from the solution of the following equation @xmath71 = 0.\label{eq_solution}\end{aligned}\ ] ] in order to see the role of nodal qps , we first ignore the fermion contribution in @xmath62 $ ] by assuming @xmath72 . in this case , the integration over @xmath73 in @xmath62 $ ] can be exactly performed , yielding @xmath63 = \frac{1}{4\pi}(\lambda-\sqrt{\gamma}\psi).\end{aligned}\ ] ] the renormalized potential becomes @xmath67 = v_0[\psi ] + \frac{\gamma \lambda}{2\pi}\psi^2 - \frac{\gamma\sqrt{\gamma}}{6\pi}\psi^3.\end{aligned}\ ] ] the cubic term induced by nematic order turns the sc transition to first order @xcite , with critical point being @xmath74 .
however , since the gapless nodal qps are present even at the lowest energy , the polarization function @xmath46 should be included in the effective action @xmath75 . after including the polarization @xmath46 into @xmath62 $ ]
, integration over @xmath73 can not be carried out analytically , and numerical method will be used . after numerically solving eq .
( [ eq_solution ] ) , we found a critical value @xmath76 for the competitive interaction between sc and nematic orders .
the sc transition remains continuous if @xmath77 , and is driven to first order when @xmath78 .
we now discuss the effects of the competing nematic order on the zero-@xmath11 superfluid density .
when @xmath23 is zero or very small , the nematic order does not change @xmath70 , which is expected and trivial . as @xmath23 increases continuously , @xmath70 and @xmath68 both decrease rapidly . to estimate this effect more quantitatively
, we assume an ultraviolet cutoff @xmath79 , and choose @xmath80 . moreover , we consider a representative bare velocity ratio @xmath81 , which is an appropriate value for yba@xmath82cu@xmath83o@xmath84 @xcite .
we further choose @xmath85 , which corresponds to @xmath86k .
it is now useful to introduce a parameter given by @xmath87 with @xmath88 , and define a dimensionless coupling constant @xmath89 .
[ fig : rho0rg](a ) shows that @xmath68 is strongly suppressed by the nematic order and is completely destroyed when @xmath23 is large enough .
it is also obvious that bare velocity ratio plays a vital role : a larger anisotropy causes smaller drop of @xmath90 .
this property provides further evidence that nodal qps can not be simply ignored . for four different values of bare ratio @xmath91 ; ( b ) @xmath92 .
the open circles _ etc . _ are guides to eyes.,width=307 ]
we then turn to calculate the density of normal nodal qps , @xmath93 .
we first briefly outline the phenomenological approach of lee and wen @xcite and then generalize it to the case with critical nematic fluctuation . in the scenario of lee and wen ,
a crucial problem is how to assess the roles played by nodal and antinodal qps in the destruction of sc condensate . as revealed clearly by extensive experiments @xcite , a pseudogap phase exists above @xmath2 .
the pseudogap turns out to have the same @xmath94-wave symmetry as the sc gap @xcite , which implies the antinodal qps remain gapped as temperature increases across @xmath2 .
it was argued @xcite that the sc state is primarily destroyed by the thermal proliferation of the low lying , gapless nodal qps , and the antinodal qps are only spectators .
lee and wen further assume that the nodal qps do not carry superflow @xcite , so the fermion spectrum is shifted by the vector potential to @xmath95 , where the qps energy is linearized as @xmath96 near nodes @xmath97 and @xmath98 is the normal state velocity .
the electric current is @xmath99 , where the superfluid tensor @xmath100 can be written as @xmath101 .
the zero-@xmath11 superfluid density is @xmath9 , and the normal qps density is derived from the free energy @xcite @xmath102 by the formula @xmath103 . in the non - interacting case ,
the fermion velocities are constants , so one gets @xmath104 which exhibits a linear temperature dependence , in agreement with experiments @xcite . at @xmath0 , @xmath1 takes its maximum value , @xmath90 .
as @xmath11 is increasing from zero , @xmath1 decreases rapidly due to thermally excited nodal qps and eventually vanishes as @xmath105 . by taking @xmath106 , it is easy to obtain @xmath107 this linear doping dependence of @xmath2 is well consistent with the uemura plot .
the same results were later reproduced by means of green s function method @xcite .
the fermi liquid ( fl ) corrections to these results were also investigated @xcite .
an important fact is that both @xmath1 and @xmath2 contain the velocity ratio @xmath108 , so these two quantities are expected to be significantly affected by the critical nematic fluctuation which can cause a nontrivial renormalization of the velocity ratio .
the above approach of computing @xmath2 is applicable when the nodal qps are well defined . at the nematic qcp @xmath3 ,
the critical fluctuation of nematic order couples strongly to gapless nodal qps and may lead to breakdown of fl behavior . as a consequence , the nodal qps might no longer be well defined qps . in a strict sense , the validity of the simple @xmath94-wave bcs theory @xcite and its fl - interaction generalizations @xcite are both doubtful . in order to proceed , here we assume that the basic approach of ref .
@xcite is still valid after the free energy of qps receives a singular correction due to nematic fluctuation .
the free fermion propagator shown in eq .
( [ eq_g_0 ] ) is renormalized by nematic order to @xmath109 where the self - energy is @xmath110 the pole in @xmath111 determines the renormalized energy @xmath112 , where @xmath113 is given by pole of free propagator @xmath114 and @xmath115 represents the interaction correction to energy . to describe this renormalization procedure , it is convenient to expand the self - energy formally as @xmath116 which exhibits unusual logarithmic behaviors since @xmath117 contains such a term as @xmath118 .
we combine this self - energy with @xmath119 , and find that the bare velocities acquire strong @xmath120-dependence , @xmath121 .
@xmath122 are very complicated functions of @xmath120 , and thus can not be written down analytically .
after straightforward numerical computation , we show the @xmath120-dependence of fermion velocities and their ratio in fig .
( [ fig : v ] ) , for the representative bare value @xmath123 .
it is clear that both @xmath124 and @xmath125 vanish as @xmath126 .
nevertheless , @xmath124 approaches zero much more slowly than @xmath125 does .
therefore , the velocity ratio vanishes at the lowest energy , @xmath127 , which recovers the extreme velocity anisotropy @xcite .
such velocity renormalization causes breakdown of fl behavior @xcite and can faithfully characterize the non - fl nature of qps at @xmath3 .
, @xmath128 , and @xmath91 with momentum @xmath120 for bare ratio @xmath129 . as @xmath126 ,
both velocities are reduced down to zero .
however , @xmath128 decreases much faster than @xmath6 , so velocity ratio @xmath130.,width=288 ] although the nodal qps are strongly damped by the nematic fluctuations and thus no longer well - defined at @xmath3 , it is still reasonable to assume that they do not carry superflow @xcite , which allows us to write @xmath131 , where the normal velocity @xmath98 becomes strongly @xmath120-dependent .
now the new free energy is @xmath132 from which we obtain a renormalized superfluid density @xmath133 at nematic qcp , @xmath2 is changed from its original value to a renormalized value @xmath134 . when @xmath135 , the renormalized superfluid density vanishes , @xmath136 , so that @xmath137 which builds a relationship between @xmath138 and @xmath139 . the ratios of @xmath140 and @xmath92 can be numerically calculated according to the above three equations . to include the influence of interaction corrections , the velocities @xmath141 appearing in @xmath68
are also replaced by @xmath122 .
the numerical results of @xmath92 are shown in fig .
[ fig : rho0rg](b ) , and those of @xmath140 given in fig .
( [ fig : rhot ] ) .
the suppression of superconductivity is prominent even when the competitive interaction between sc and nematic orders is quite weak , and is further enhanced by this interaction . as aforementioned
, the sc phase transition at @xmath134 remains continuous as long as @xmath23 is small . .
( a ) : @xmath142 ; ( b ) : @xmath143.,width=307 ] clearly , the extent to which @xmath1 and @xmath2 are suppressed by the nematic order depends on a number of parameters , such as @xmath25 , @xmath23 , and @xmath91 .
it is helpful to make more quantitative analysis .
the magnitudes of @xmath25 as well as @xmath144 have already been chosen .
the bare velocity ratio is still fixed at @xmath129 .
[ fig : rho0rg](b ) shows that @xmath2 is roughly suppressed by @xmath145 when @xmath142 , and by @xmath146 when @xmath147 .
the drop of @xmath18 can be analyzed in a similar way .
an important implication of fig .
( [ fig : rhot ] ) is that @xmath18 is more strongly reduced at higher temperatures .
experimentally , the suppression of @xmath2 can be tested by measuring resistivity and meissner effect , and the suppression of superfluid density may be probed by measuring london penetration length @xmath148 , preferably using microwave techniques @xcite based on the relationship @xmath149 .
the strong suppression effects rely on the divergence of correlation length @xmath150 of the nematic field @xmath22 at the qcp @xmath3 . as one moves away from the point @xmath3 , the strong suppression of superfluid density and @xmath2
will be rapidly diminished .
in fact , once correlation length @xmath150 becomes finite , the interaction between nematic fluctuation and nodal qps is no longer singular .
consequently , the self - energy of nodal qps , @xmath151 , does not display logarithmic behavior , which indicates that the fermion velocities only receive unimportant renormalizations . in this case ,
the density of normal qps is not significantly modified by the competing nematic order , since the velocity ratio does not depart far from its bare value .
the influence of nemtic order on @xmath90 is also weakened . for @xmath152 ,
the fluctuation of @xmath22 is rather weak and does not change @xmath90 much if @xmath23 is not very large . for @xmath153 ,
the fluctuation of @xmath22 is gapped and also leads to much smaller change of @xmath90 compared with what happens at nematic qcp @xmath3 .
therefore , the @xmath2 curve generally follows the uemura plot in the whole underdoped region , but deviates from the linear behavior in the close vicinity of @xmath3 .
since the suppression of superconductivity is most pronounced at @xmath3 , it may help to find the position of the predicted nematic qcp if such a point really exists .
in summary , we have considered the effects of nematic fluctuation on superfluid density and critical temperature @xmath2 in the vicinity of nematic qcp in the contexts of some htscs .
on one hand , the critical fluctuation of nematic order parameter reduces zero-@xmath11 superfluid density as a result of ordering competition .
on the other hand , it couples strongly to the gapless nodal qps and excites more normal qps out of the sc condensate by triggering an extreme fermion velocity anisotropy .
both of these two effects combine to significantly suppress superconductivity .
therefore , we have predicted a dip shape reduction of @xmath2 at nematic qcp @xmath3 , which is schematically shown in fig .
( [ fig:1 ] ) .
the dip shape of @xmath2 reminds us of @xmath154 anomaly @xcite , which refers to an anomalous suppression of superconductivity at doping @xmath155 . a sudden drop of @xmath90
was also observed @xcite in la@xmath156sr@xmath157cuo@xmath158 ( lsco ) at @xmath155 .
it appears that @xmath1 and @xmath2 near the nematic qcp bear some similarity to the basic features of these experiments .
however , we refrain from identifying the @xmath154 doping as the anticipated nematic qcp for several reasons : i ) @xmath154 anomaly is usually attributed to the formation of static stripe order @xcite ; ii ) it is not clear why nematic qcp exists precisely at @xmath159 , not elsewhere ; iii ) low-@xmath11 dc thermal conductivity @xmath160 was predicted to be enhanced at nematic qcp @xcite , whereas an early transport measurement found a drop of @xmath160 in lsco at @xmath159 @xcite .
recently the critical nematic fluctuations are shown to induce unusual behaviors in several quantities , such as qps spectral function @xcite , specific heat @xcite , nuclear relaxation rate @xcite , thermal conductivity @xcite , and qp interference @xcite .
normally , a single observation alone is not able to uniquely fix the nematic qcp .
should all or most of these predictions , including the suppression of superconductivity , be observed , the nematic qcp might be determined .
however , the absence of a sharp nematic qcp does not necessarily mean the absence of nematic phase , since the sharp qcp may be rounded by disorders and become a crossover @xcite . in case
this smearing occurs , many of the anomalous critical behaviors are weakened , or even destroyed , and the nematic phase most possibly shows its existence in the @xmath161 regime . in the calculations presented in this paper
, we have assumed that the sc order parameter @xmath21 is a classical field and neglected its quantum fluctuation .
this assumption is expected to be a good one if the nematic qcp @xmath3 is deep inside the sc dome , namely @xmath3 is not close to the sc critical point @xmath30 .
when @xmath3 is close to @xmath30 , the quantum fluctuation of sc order @xmath21 may become very important and a fully quantum - mechanical treatment will be necessary .
g.z.l . acknowledges support by the national natural science foundation of china under grant no .
11074234 and the visitors program of mpipks at dresden .
y. j. uemura , g. m. luke , b. j. sternlieb , j. h. brewer , j. f. carolan , w. n. hardy , r. kadono , j. r. kempton , r. f. kiefl , s. r. kreitzman , p. mulhern , t. m. riseman , d. ll .
williams , b. x. yang , s. uchida , h. takagi , j. gopalakrishnan , a. w. sleight , m. a. subramanian , c. l. chien , m. z. cieplak , gang xiao , v. y. lee , b. w. statt , c. e. stronach , w. j. kossler , and x. h. yu , phys .
* 62 * , 2317 ( 1989 ) .
e. demler , s. sachdev , and y. zhang , phys .
lett . * 87 * , 067202 ( 2001 ) ; s. a. kivelson , d .- h .
lee , e. fradkin , and v. oganesyan , phys . rev .
b * 66 * , 144516 ( 2002 ) ; r. m. fernandes and j. schmalian , phys . rev .
b * 82 * , 014521 ( 2010 ) ; e. g. moon and s. sachdev , phys . rev .
b * 82 * , 104516 ( 2010 ) ; d. chowdhury , e. berg , and s. sachdev , phys .
b * 84 * , 205113 ( 2011 ) .
j. gonzalez , f. guinea , and m. a. h. vozmediano , nucl .
b * 424 * , 595 ( 1994 ) ; d. h. kim , p. a. lee , and x .-
wen , phys .
rev . lett . * 79 * , 2109 ( 1997 ) ; o. vafek , phys . rev . lett . * 98 * , 216401 ( 2007 ) . | a nematic quantum critical point is anticipated to exist in the superconducting dome of some high - temperature superconductors .
the nematic order competes with the superconducting order and hence reduces the superconducting condensate at @xmath0 .
moreover , the critical fluctuations of nematic order can excite more nodal quasiparticles out of the condensate .
we address these two effects within an effective field theory and show that superfluid density @xmath1 and superconducting temperature @xmath2 are both suppressed strongly by the critical fluctuations .
the strong suppression of superconductivity provides a possible way to determine the nematic quantum critical point . |
the lyman break ` dropout ' technique was first applied to select lyman break galaxies ( lbgs ) at @xmath14 @xcite , and since then it has been extensively used to select lbg candidates at @xmath15 ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this dropout technique has generated large samples of star - bursting galaxy candidates at @xmath15 , but there is only one major study @xcite that investigates lbgs at @xmath0 based on dropout selection criteria .
the primary reason for this is that we need highly sensitive space - based cameras to observe the mid- to near - ultraviolet ( uv ) wavelengths required to select lbgs at @xmath0 .
the new wide field camera 3 ( wfc3 ) on the refurbished hubble space telescope ( hst ) with its superior sensitivity compared to the wide - field planetary camera 2 ( wfpc2 ) or the galaxy evolution explorer ( galex ) and filters below the atmospheric cut - off wavelength ( e.g. , f225w and f275w ) , allows us to photometrically identify and study lower redshift ( @xmath0 ) lbgs .
the improved sensitivity / depth allows us to probe the lower luminosity systems at these redshifts .
there are two important reasons to understand these lbgs .
first , to study the star formation properties of these lbgs , because they are at redshifts corresponding to the peak epoch of the global star formation rate ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and , secondly , they are likely lower redshift analogs of the high redshift lbgs because of the similar dropout selection at all redshifts whose understanding will help shed light on the process of reionization in the early universe ( e.g. , * ? ? ?
* ; * ? ? ?
the major advantage of identifying and studying various properties including star formation properties of lower redshift lbgs is that these lbgs can be investigated in rest - frame uv _ as well as _ rest - frame optical filters .
the high redshift lbgs have very little information on their rest - frame _ optical _ properties , so a detailed understanding of lower redshift lbgs is very important to get insight into the physical and morphological nature of high redshift lbgs . the new uv observations of the wfc3 science oversight committee ( soc ) early release science extragalactic program ( pid : 11359 , pi : oconnell ; hereafter `` ers2 '' ) , covers approximately 50 arcmin@xmath1 in the north - western part of the great observatories origins deep survey ( goods ; * ? ? ?
* ) south field .
here we use the high sensitivity of the new wfc3 uvis channel data , along with existing deep optical data obtained with the advanced camera for surveys ( acs ) as part of the goods program , to search for lbg candidates at @xmath0 .
we use dropout color selection criteria based on color - color plots , obtained with the wfc3 uvis and acs filters to find three unique sets of uv dropouts f225w - dropouts , f275w - dropouts and f336w - dropouts which are lbg candidates at @xmath2 , @xmath3 and @xmath4 , respectively ( as shown in figure [ fig : lybreak ] ) .
this paper is organized as follows : in [ data ] we summarize the wfc3 ers2 observations , and in [ sample ]
we discuss the selection , and in [ reliable ] the reliability of our color selected @xmath0 lbg sample . in [ results ]
we discuss the data analysis , which includes measuring their number counts and surface density ( [ ncounts ] ) , and compare these with other surveys at higher redshifts , and estimate rest - frame uv luminosity functions ( [ lfs ] ) for these samples . in [ conclusion ] we conclude with a summary of our results . in the remaining sections of this paper we refer to the hst / wfc3 f225w , f275w , f336w , filters as @xmath16 , @xmath17 , @xmath18 , and hst / acs f435w , f606w , f775w , f850lp filters as @xmath19 , @xmath20 , @xmath21 , @xmath22 , respectively , for convenience .
we assume a _
wilkinson microwave anisotropy probe _
( wmap ) cosmology with @xmath23=0.274 , @xmath24=0.726 and @xmath25=70.5 km s@xmath26 mpc@xmath26 , in accord with the 5 year wmap estimates of @xcite .
this corresponds to a look - back time of 10.37 gyr at @xmath27 .
magnitudes are given in the ab@xmath28 system @xcite .
the wfc3 ers2 observations were done in both the uvis ( with a fov of 7.30 arcmin@xmath1 ) and the ir ( with a fov of 4.65 arcmin@xmath1 ) channels .
details of these observations are described in @xcite . here
we briefly summarize the uv imaging observations .
the wfc3 ers2 uv observations were done in three broad - band filters @xmath16 , @xmath17 and @xmath18 , whose total throughput curves are shown in figure [ fig : lybreak ] .
the ers2 field covers the north - western @xmath2950 arcmin@xmath1 of the goods - south field , and was observed in 8 pointings with a 2@xmath304 grid pattern during september - october 2009 .
the @xmath16 and @xmath17filters were observed for 2 orbits per pointing , while the @xmath18filter was observed for 1 orbit per pointing , for a total of 40 orbits over the full ers2 field .
the raw images were processed through the ` calwf3 ` task ( using the latest version as of december 1 , 2009 ) included in the stsdas package ( version 3.11 ) , and the latest reference files from the stsci .
the flat - fielded images were then aligned and drizzled using ` multidrizzle ` @xcite onto the same grid as the goods - south v2.0 acs data , which were rebinned to a pixel size of 0.09 .
the final uv image mosaics have a pixel scale of 0.09 to match the wfc3 ir image mosaics in all filters , and cover @xmath2950 arcmin@xmath1 area in the goods - south field .
the combination of the three wfc3 uv filters and the four acs optical filters provide excellent capability of selecting galaxies at @xmath0 , using the dropout technique to detect the lyman - break signature that occurs at a rest - frame wavelength of 912 @xcite .
figure [ fig : lybreak ] shows the locations of the rest - frame 912 lyman break at various redshifts .
it is clear that three wfc3 uvis filters , along with the acs @xmath19- and @xmath20-bands are very useful in identifying lbg candidates at @xmath0 .
we performed matched - aperture photometry by running the ` sextractor ` @xcite algorithm in the dual - image mode with the corresponding rms maps .
the rms maps were derived from the ` multidrizzle ` generated weight maps , following the procedure discussed in @xcite .
we have measured the 10-@xmath31 point source detection limits in a 0.2 aperture as 26.0 , 26.1 and 25.7 ab - mag in @xmath16 , @xmath17 and @xmath18 , respectively .
we constructed three separate catalogs by using the three separate images ( @xmath17 , @xmath18 , @xmath19 ) as detection images .
these catalogs are referred as the @xmath17-based , @xmath18-based and @xmath19-based catalogs .
we used the wfc3 in - flight photometric zeropoints ( 24.06 , 24.14 , 24.64 ab - mag for @xmath16 , @xmath17 & @xmath18 ; * ? ? ?
* ) obtained from stsci website .
our initial selection of uv dropouts is based on dropout color criteria obtained from the stellar population models of ( * ? ? ?
the top three panels of figure [ fig : clr - clr ] shows the bc03 star - forming galaxy models for three dropout samples with @xmath32 mag ( solid black lines ) , expected colors of stars ( black dots ) from @xcite , and tracks of low - redshift ellipticals ( gray lines ) from @xcite and @xcite .
the gray shaded region in top panels is our selection region .
though galactic stars clearly land in the selection region in figure [ fig : clr - clr ] , these are easily removed by simple morphological criteria , as in @xcite .
we have applied the @xcite prescription to estimate igm attenuation for proper comparison with other studies .
we also checked how bc03 tracks are affected by applying the madau prescription only to galaxies at @xmath33 ( i.e. no igm attenuation below @xmath34 ) , and only to galaxies at @xmath35 , to see the effects of fluctuations in igm attenuation .
we find that our adopted selection criteria will still be able to pick - up star - forming galaxies with @xmath36 .
the selection criteria adopted here are similar to the criteria used to identify lbg candidates at @xmath37 ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we use @xmath17-based catalogs to select @xmath16-dropouts using ( @xmath16 @xmath17 ) vs. ( @xmath17
@xmath18 ) color - color space , as shown in the bottom leftmost panel of figure [ fig : clr - clr ] . for @xmath16-band dropouts ,
we require : @xmath38 \:\hbox{mag } \\ \hbox{and}\ : ( u_{\rm 336}-b_{\rm 435 } ) > -0.5 , \ : [ s / n(u_{\rm 275 } ) ] > 3,\ : [ s / n(u_{\rm 225 } ) ] < 3 \end{array } \right.\ ] ] here , the @xmath39 is defined as 1.0857 divided by the ` sextractor ` error in the total magnitude .
the ` sextractor ` magnitude uncertainties are estimated from carefully constructed rms maps that account for correlated pixel noise and hence , gives better estimate of the noise .
details of the wfc3 data reduction process are given in @xcite .
we require ( @xmath16
@xmath17 ) @xmath40 1.3 mag , which is redder than what we have applied for other dropouts , because there is no bluer filter available than @xmath16 to confirm that these dropouts are undetected at wavelengths bluer than rest - frame 912 .
we have also applied the additional criterion ( @xmath18 @xmath19 ) @xmath40 0.5 mag to eliminate the possibility of selecting spurious candidates , since this color - color space is based only on three uv filters , and it is required that lbg candidates be detected and are bright enough in the @xmath19-band .
the @xmath19-band is about @xmath2922.5 mag deeper than the @xmath18-band , so a simple @xmath39 ( @xmath19 ) @xmath40 3 cut can not be used , because we would still pick - up very faint objects in @xmath19-band .
there are 106 objects inside the selection region .
we find a total of 70 lbg candidates ( @xmath16-dropouts ) based on these selection criteria .
after visually checking each candidate using the 10-band ( 3 wfc3 uv , 4 acs optical and 3 wfc3 ir ) hst imaging from the ers2 goods observations , we eliminated 4 candidates from our sample as spurious ( due to their closeness to a bright object or a probable faint stellar diffraction spike ) .
this examination leaves us with 66 @xmath16-dropouts .
similarly , we use @xmath18-based catalogs to select @xmath17-dropouts using ( @xmath17 @xmath18 ) vs. ( @xmath18
@xmath19 ) color - color space as shown in the bottom middle panel of figure [ fig : clr - clr ] . for @xmath17-band dropouts ,
we require : @xmath41 \:\hbox{mag } \\ \hbox{and}\ : [ s / n(u_{\rm 336 } ) ] > 3,\ : [ s / n(u_{\rm 275 } ) ]
< 3 , \ : [ s / n(u_{\rm 225 } ) ] < 1 \end{array } \right.\ ] ] there are 223 objects inside the selection region .
we find a total of 153 lbg candidates ( @xmath17-dropouts ) based on these selection criteria .
after visual examination , we eliminated 2 candidates from our sample as spurious , because of their proximity to a brighter object . therefore , our core sample contains 151 @xmath17-dropouts .
finally , we use @xmath19-based catalogs to select @xmath18-dropouts using ( @xmath18 @xmath19 ) vs. ( @xmath19
@xmath20 ) color - color space as shown in the bottom rightmost panel of figure [ fig : clr - clr ] . for @xmath18-band dropouts ,
the following color selection was applied : @xmath42 \:\hbox{mag } \\ \hbox{and}\ : [ s / n(b_{\rm 435 } ) ] > 3,\ : [ s / n(u_{\rm 336 } ) ] < 3,\ : [ s / n(u_{\rm 275 } ) ] < 1 , \ : [ s / n(u_{\rm 225 } ) ]
< 1 \end{array } \right.\ ] ] we require ( @xmath18
@xmath19 ) @xmath40 0.8 mag , which is bluer than what we have applied for other dropouts , because we have two bands bluer than @xmath18 to confirm that these dropouts are not detected ( @xmath43 ) at wavelengths bluer than rest - frame 912 .
there are 1156 objects inside the selection region .
we find a total of 260 lbg candidates ( @xmath18-dropouts ) based on the above mentioned selection criteria .
after visually checking each candidate , we eliminated 4 spurious candidates ( same reason as discussed before for other dropout candidates ) .
the final sample consists of 256 @xmath18-dropouts .
the gray data points in the bottom panels of figure [ fig : clr - clr ] shows _ all _ the sources in our catalogs that are not selected as lbg candidates .
those that fall within our selection regions were excluded from our candidate samples mainly because of our low @xmath39 cuts ( @xmath44 ) in the bluer bands and the hard magnitude limit ( @xmath45 mag ) in the selection band .
the primary reason for this is the varying depth of these filters .
the @xmath19 is much deeper than @xmath18 , so when we select @xmath18-dropouts with @xmath19 @xmath45 mag , we are not selecting galaxies fainter than 26.5 mag in @xmath19 .
we can not conclusively say whether these faint galaxies are lbg candidates or not because of the shallower @xmath18 images . similarly , @xmath16 and @xmath17 are slightly deeper than @xmath18 , so while selecting the @xmath17- or @xmath18-dropouts , our selection criteria of @xmath46 in the bluer bands still keeps some faint objects ( with @xmath47 ) in the selection region which are not selected as dropouts .
therefore , the varying depth between filters , combined with our magnitude and @xmath39 cuts , are responsible for objects in the selection region that are not selected as lbg candidates .
overall , our selection is conservative , because of these constraints .
there are a few compact objects in our selected samples , but when we visually check these objects in the 10-band hst imaging , and compare with the more robustly selected star catalog of @xcite , we can not confirm any stars .
@xcite gives a detailed discussion of the star - galaxy separation procedure used for the ers2 data , and confirms that within our sample magnitude range ( 2426.5 mag ) there are practically no stars in the uv bands , and very few in the b - band .
there could be a weak agn in some of these galaxies , which we will investigate in our future paper on stellar populations and spectral analysis of these lbgs .
figure [ fig : image ] shows three example images of color selected uv - dropouts whose redshifts are confirmed by ground - based spectroscopy ( see next section ) .
these examples are shown here in the 10-band hst imaging obtained from ers2 observations .
the final sample consists of 66 @xmath16- , 151 @xmath17- and 256 @xmath18-dropouts .
there is one object overlapping between @xmath16- and @xmath17-dropout samples , while seven objects are in common between @xmath17- and @xmath18-dropout samples . in order to test reliability of our color selection ,
we compare our dropout samples with spectroscopic redshifts from the very large telescope ( vlt ; e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and with the 10-band ( 3 wfc3 uv , 4 acs optical and 3 wfc3 ir ) photometric redshifts obtained from our ers2 observations ( cohen et al .
2010 , in prep ) .
when we match our dropout catalogs with these redshift catalogs , we find that @xmath2980% of our dropouts have photometric redshifts , but only @xmath2930% have spectroscopic redshifts .
though photometric redshifts are from the same ers2 dataset , we do nt have 100% matching catalogs .
the ers2 photometric redshifts are based on the @xmath48-band selected catalogs , and the wfc3 ir channel covers a smaller area than the wfc3 uvis or acs / wfc cameras , which were used here to identify these uv dropouts .
the most likely reason for the low number of spectroscopic confirmations is the ` redshift desert ' .
the galaxies in this redshift range ( @xmath49 ) are difficult to identify via ground - based spectroscopy , because of the lack of strong features in 4500 - 9000 range , where most spectrographs on large telescopes are optimized .
figure [ fig : redshifts ] shows the redshift distributions of the dropout lbg candidates three dropout samples are shown in three separate panels with spectroscopic and photometric redshifts .
the hashed ( solid gray ) histogram and solid ( dot - dash ) curve shows the distribution and the best gaussian fit to the number of lbg candidates with photometric ( spectroscopic ) redshifts .
table [ tab : redshifts ] lists the number of dropouts with spectroscopic / photometric redshifts and their average redshifts obtained from the distributions in figure [ fig : redshifts ] .
figure [ fig : redshifts ] shows that based on these available redshifts , our dropout selection is very reliable , and that wfc3 uv filters provide a very efficient way to select lbgs at @xmath0 .
spectroscopic redshifts are only available for @xmath2930% of our dropout galaxies , and have on average @xmath296% outliers or low-@xmath50 interlopers ( @xmath295% for @xmath16- , @xmath295% for @xmath17- and @xmath299% for @xmath18-dropouts ) . the outliers are defined as any object at @xmath51 . the ers2 photometric redshifts are available for @xmath2980% of our dropouts and have on average @xmath2912% outliers with the most ( @xmath2915% ) amongst the @xmath16-dropouts , as expected because of the lack of any available wfc3 bands bluer than @xmath16 .
the comparison of our color selected dropouts with spectroscopic and photometric redshifts show that the fraction of outliers in both cases are comparable or better than the fraction of outliers in spectroscopic follow - up surveys of star - forming galaxies and dropout selected lbgs at @xmath52 , which is about @xmath295 - 15% @xcite . a similar comparison with publicly available photometric redshifts ( e.g. , * ? ? ?
* ; * ? ? ?
* ) shows that they have a higher percentage of outliers ( @xmath2917% ) , because these large surveys have mostly shallow ground - based near - uv data which can not go bluer than @xmath293000 , because of the atmospheric cut - off .
therefore , space - based wfc3 uv data are essential to get accurate photometric redshifts for these lower redshift galaxies ( @xmath0 ) .
the average spectroscopic or photometric redshift for @xmath18-dropouts from figure [ fig : redshifts ] is @xmath53 , but based on the location of the lyman break ( figure [ fig : lybreak ] ) , the average redshift for @xmath18-dropouts should be about @xmath54 .
we have identified small number of brighter ( @xmath55 mag ) lbg candidates at @xmath56 , but we are missing a significant number of fainter ( @xmath57 mag ) lbg candidates at these redshifts .
the main reason is that we require redder colors ( @xmath18
@xmath19 @xmath58 2.0 mag ) to select higher redshift galaxies in this dropout sample , because the ly@xmath10 forest absorption at 912 - 1216 begins to increasingly affect the @xmath18 band at these redshifts .
this implies that we need the @xmath18 images to be @xmath2911.5 mag deeper to consistently select all dropouts at @xmath59 , improve the photometric redshift distribution , and lower the number of outliers for this sample .
therefore , we have a relatively smaller number of lbg candidates ( within our magnitude limit ) at @xmath56 in the @xmath18-dropouts sample .
the observed raw number counts of lbg candidates at @xmath0 at a rest - frame wavelength of 1700 are shown in figure [ fig : ncounts ] .
when we combine all three dropout samples , the average photometric redshift is @xmath8 . for proper comparison ,
these number counts are _ not _ corrected for incompleteness or cosmic variance , and therefore , the counts start to drop at fainter magnitudes ( @xmath60 mag ) .
figure [ fig : ncounts ] ( top panel ) shows number counts ( in number per arcmin@xmath1 per 0.5 mag bin ) of _ all _ dropouts ( @xmath0 ) in our sample compared with other ground - based and space - based lbg surveys @xcite at @xmath61 .
we have also plotted @xmath62 number counts from @xcite to show the change in the surface densities ( number per arcmin@xmath1 ) as a function of redshift .
* @xmath14 ) used ground - based imaging in @xmath2914 fields , with each field observed for many kilo - seconds ( ks ) , followed by ground - based spectroscopy to confirm many of their color selected candidates .
the @xcite selection was based on lbg color criteria down to ab@xmath2925 mag .
@xcite observed the goods - south field using vlt / vimos to get deep @xmath63-band imaging ( ab@xmath2927 mag ) .
their number counts for lbg candidates at @xmath14 shown in figure [ fig : ncounts ] come from the deepest part of the vimos field , which covers @xmath2988 arcmin@xmath1 with exposure time of @xmath2920 hours ( 72 ks ) . on the other hand
, @xcite observed the subaru deep field @xcite using deep ( @xmath40100 ks ) near - uv imaging from the space - based galex observations ( with @xmath295 fwhm resolution ) to select lbg candidates at @xmath8 down to ab@xmath2925 mag , and used ground - based spectroscopy to confirm many lbgs at @xmath8 . from figure
[ fig : ncounts ] ( top panel ) we note three major points .
first , there is only one space - based galex lbg survey at @xmath8 @xcite , which clearly shows that the wfc3 uv observations with better sensitivity and resolution can play a vital role in identifying lbgs at @xmath64 .
secondly , all surveys mentioned above use deep uv imaging with @xmath65 ks exposures , while our wfc3 uv observations are only @xmath66 ks ( 1 to 2 orbits ) , and still we find that our observations are @xmath290.51.0 mag deeper compared to some of these surveys .
finally , our numbers agree very well with the decreasing trend of lbg surface densities as a function of redshift from @xmath5 to @xmath67 , which we will address quantitatively in the next section .
the bottom panel of figure [ fig : ncounts ] shows number counts for each dropout sample .
the @xmath17- and @xmath18-dropout samples show comparable number counts and agree generally with surveys at higher redshifts , but the @xmath16-dropouts show lower number counts . given the trend with redshift shown by other samples in the upper panel of figure [ fig : ncounts ]
, we would expect more @xmath16-dropouts than other dropouts at higher redshifts .
the numbers are smaller than expected because of the conservative selection criteria we applied owing to the absence of a second filter below the lyman break ( see [ sample ] ) to confirm our dropout selection .
this approach led us to small numbers of @xmath16-dropouts in a relatively narrow redshift range around @xmath2 .
hence , we do nt have a fully representative @xmath16-dropout sample , but with the future deeper observations we can use a somewhat more liberal selection criteria to get better statistics for this sample .
we calculated the rest - frame uv luminosity functions ( lf ) using the @xmath68 method ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) in 0.5 mag wide bins .
the absolute magnitudes of lbg candidates were measured in the observed bands that are equivalent to rest - frame 1500 to minimizes @xmath69-corrections , and using the average redshift for each object in each sample ( @xmath2 , @xmath3 , @xmath4 , respectively ) .
these absolute magnitudes are uncorrected for internal dust absorption .
we compute lfs for the three dropout samples : @xmath16-dropouts ( @xmath2 lbg candidates ) , @xmath17-dropouts ( @xmath70 lbg candidates ) , and @xmath18-dropouts ( @xmath71 lbg candidates ) .
figure [ fig : lf ] shows the lfs for these three dropout samples .
we model these lfs with a standard schechter function @xcite , which is parametrized by the characteristic absolute magnitude ( @xmath72 ) , the normalization ( @xmath73 ) , and the faint - end slope ( @xmath10 ) .
the shaded gray regions in figure [ fig : lf ] show the uncertainty in the lf based on 1-@xmath31 uncertainty in @xmath9 and @xmath10 .
table [ tab : lfs ] lists the best - fit schechter function parameters @xmath9 , @xmath10 and @xmath74 for these three dropout samples . to investigate incompleteness in each redshift bin , we ran simulations to calculate @xmath75 , which is the probability that a galaxy of apparent magnitude @xmath76 and at redshift @xmath50 will be detected in the image _ and _ will meet our color selection criteria . in these simulations ,
large numbers of artificial objects with a range of redshifts and magnitudes were added to the real ers2 images , and then recovered using exactly the same method and selection criteria that were employed for the real observations .
for these simulations , we used bc03 models assuming salpeter initial mass function ( imf ) , constant sfr , solar metallicity , @xmath77 , an age of 1 gyr with different redshift range for each sample and varying magnitudes .
these models were used to generate color and extinction properties of our artificial objects .
we chose artificial objects to be point - like sources .
the selection function obtained from this exercise ( adding and recovering artificial objects ) is similar in shape as the distributions in figure [ fig : redshifts ] , and the mean redshift value obtained from these simulations ( for each sample ) is within 1-sigma of the mean value obtained in figure [ fig : redshifts ] . these @xmath75 estimates were used to determine @xmath68 for the lf .
we did not make the corrections for interlopers in our lf estimates .
there are five main reasons for this .
first , we have checked our lfs by boosting the errors by 10% , and we find that the best fit values remain the same , while the uncertainties on these values increases slightly .
second , the limited number of spectroscopic redshifts ( @xmath78% candidates have spectroscopic redshifts ) does not give us a correct estimate of interlopers in our color selected sample .
third , the total fraction of spectroscopic interlopers is very small ( @xmath79% ) , and when we subdivide them as a function of magnitude it is even smaller .
fourth , the estimate of interlopers based on photometric redshifts is not very accurate because of uncertainty in the photometric redshifts . though the ers2 photometric redshifts ( cohen et al .
2010 , in prep ) are better than some of the publicly available photometric redshifts , they are still uncertain by a few percent .
finally , for the @xmath17- and @xmath18-dropouts the faintest bin is most affected by the interlopers , but that data point is already uncertain because of very few objects in that bin .
the leftmost panel of figure [ fig : lf ] shows the resulting lf for @xmath16-dropouts .
the three brightest bins contain on average 3 objects per bin , and hence they are more uncertain . that leaves us with only three data points with a statistically significant number of objects .
it is not possible to fit a schechter function to three data points by keeping all three parameters free . in the absence of deeper data
, we fix the faint - end slope , @xmath10 , based on the best - fit observed trend between redshifts and @xmath10 for lbgs at @xmath80 ( figure [ fig : mstr ] ) .
the best - fit parameters for this dropout sample are meant to be mostly illustrative due to low number statistics .
the middle panel of figure [ fig : lf ] shows the lf for the @xmath17-dropouts and the rightmost panel shows the lf for the @xmath18-dropouts .
it is difficult to estimate the faint - end slope from these observations , as we can see from the uncertain faintest point in the lfs of @xmath17- and the @xmath18-dropouts .
though our best - fit estimates are very close to what we expect at these redshifts ( @xmath0 ) from other studies ( e.g. , * ? ?
* ; * ? ? ?
* ) at nearby redshifts , we will need deeper ( @xmath2912 mag ) uv observations to properly constrain the faint - end slope for these three dropout samples . in general
, it is not straightforward to directly compare our lfs with those from previous studies .
first , our redshift range is different , and this is the first time that this camera and filter set have been used to select lbgs .
secondly , in some cases the adopted cosmologies are slightly different .
it is well known ( e.g. , * ? ? ?
* ) that the derived lfs strongly depends on the assumed cosmological models , but the evolutionary trends seen in the lfs in our three redshift bins are virtually independent of the assumed cosmology .
figure [ fig : mstr ] shows the evolutionary trends in our three redshift bins , as well as comparisons to other studies on lbgs at different redshifts .
the top panel of figure [ fig : mstr ] shows the faint - end lf slope , @xmath10 , as function of redshift .
the @xcite @xmath81 sample is based on the spectroscopically confirmed galaxies with the galex near - uv detection ( @xmath82 mag ) , and the @xmath83 sample is based on the photometric redshifts .
the @xcite samples are not selected based on lyman break color criteria but because of the lack of lbg candidates at @xmath84 , we have used this star - forming galaxies sample for comparison .
the @xcite and the @xcite samples are dropout selected lbgs at @xmath14 and @xmath8 , respectively .
the black line is the best - fit observed trend between @xmath10 and @xmath50 for lbgs at @xmath80 , which is very similar to that of @xcite .
the observed trend is that as the redshift increases , the faint - end slope , @xmath10 , becomes steeper ( more negative ) , illustrating that lower luminosity dwarf galaxies dominate the galaxy population at higher redshifts .
excluding the fixed @xmath10 data point at @xmath2 , our data points at @xmath70 and @xmath71 agree very well within the current uncertainties with the black line , as well as with other data points in close redshift proximity .
the bottom panel of figure [ fig : mstr ] shows the characteristic absolute magnitude , @xmath9 , as a function of redshift .
again , the general observed trend is that as redshift increases , the characteristic absolute magnitude , @xmath9 , becomes brighter ( more negative ) until @xmath85 .
this trend is considered as an evidence of ` downsizing ' galaxy formation scenario ( e.g. , * ? ? ?
* ) , where luminous massive galaxies form at higher redshifts .
our first data point at @xmath2 follows this general trend , but it is more uncertain due to the limited statistics in this dropout sample .
the other two data points at @xmath70 and @xmath4 fit very well within the evolutionary trend seen at these and higher redshifts .
figure [ fig : mstr ] shows rapid decline of @xmath9 between @xmath14 and extending to @xmath86 .
this turnover is well defined in our and @xcite samples .
it is important for future surveys to exploit the special capabilities of the wfc3 in the near - uv to obtain larger samples to understand the relation between this critical transition in @xmath9 and physical processes in lbgs at @xmath87 .
@xcite have used deep ground - based imaging data to constrain the uv lf of the ` bx ' ( e.g. , * ? ? ?
* ) population at @xmath88 , which selects star - forming galaxies based on @xmath89 colors .
when we compare our lfs with that of the ` bx ' population , we find some differences in @xmath9 and @xmath10 values .
first , our @xmath16-dropout sample has lower redshift ( @xmath90 ) compared to the ` bx ' population ( @xmath91 ) , so our @xmath9 value ( 19.43 mag ) is fainter than their value of 20.70 mag , and agrees with the general trend discussed above .
second , @xmath9 values of our @xmath17-dropout sample and the ` bx ' sample agree within our 1-@xmath31 uncertainty , while the @xmath10 is little steeper for the ` bx ' sample .
we believe that complete agreement between our lbgs sample and the ` bx ' sample is not possible , because although the ` bx ' selection selects star - forming galaxies , it is very likely that the dropout selected sample at a similar redshift might not be same as the ` bx ' selected sample .
some galaxies which are selected through the ` bx ' color selection criteria might not be in the dropout selected sample , and vice versa .
therefore , it is difficult to directly compare the ` bx ' and the lbg samples , and the differences in these samples could cause the lf parameters at similar redshift to differ ( see also * ? ? ?
therefore , for both @xmath9 and @xmath10 our results agree very well with the expected observed trends ( figure [ fig : mstr ] ) as a function of redshift . at lower redshifts ( @xmath87 ) , our data points will help to reduce the gap between the well studied @xmath6 and @xmath7 regimes .
the agreement with the observed evolutionary trend of m@xmath92 and @xmath10 also show the reliability of our lfs , which can be improved with the future deeper and wider wfc3 uv observations ( e.g. , candels multi - cycle treasury program # 12060 - 12064 , pi : s. faber ) .
we use newly acquired uv observations from the wfc3 uvis channel along with existing deep acs observations of the goods program , to identify uv dropout galaxies , which are lbg candidates at @xmath0 .
we find 66 @xmath16- , 151 @xmath17- and 256 @xmath18-dropouts to a magnitude limit of [email protected] mag .
this allows us to estimate rest - frame uv lfs in three redshift bins ( @xmath2 , @xmath3 , @xmath4 ) .
their best - fit schechter function parameters @xmath72 , @xmath10 and @xmath73 agree very well with the observed evolution of these parameters with respect to redshift .
we need space - based uv imaging to identify and understand the @xmath11 lbgs selected based on their uv dropout signature .
the new wfc3 uvis camera on the hst now allows us to do that with much better sensitivity and resolution than galex , and has opened up a new regime of detailed uv imaging studies of low to intermediate redshift ( @xmath11 ) lbgs , which is not possible from the ground due to the atmospheric cut - off .
the quality of rest - frame near - uv imaging ( @xmath583000 ) of these galaxies greatly exceeds that which can be done with ground - based near - uv observations .
future work will investigate the morphology and stellar populations of these lower redshift lbgs , to better understand their higher redshift counterparts .
the upcoming wfc3 uvis imaging surveys deep and wide have the potential to robustly measure the evolution of lbgs at @xmath11 , and with uniform selection all - the - way to very high redshifts provide better understanding of very high redshift lbgs .
we thank the referee for helpful comments and suggestions that significantly improved this paper .
we thank m. nonino , c. ly and their collaborators for providing their lbg number counts .
this paper is based on early release science observations made by the wfc3 scientific oversight committee .
we are grateful to the director of the space telescope science institute for awarding director s discretionary time for this program .
finally , we are deeply indebted to the brave astronauts of sts-125 for rejuvenating hst .
support for program # 11359 was provided by nasa through a grant from the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas 5 - 26555 .
adelberger , k. l. , steidel , c. c. , shapley , a. e. , et al .
2004 , apj , 607 , 226 arnouts , s. , schiminovich , d. , ilbert , o. , et al .
2005 , apj , 619 , l43 balestra , i. , mainieri , v. , popesso , p. , et al .
2010 , a&a , 512 , a12 bertin , e. , & arnouts , s. 1996 , a&as , 117 , 393 bouwens , r. j. , illingworth , g. d. , franx , m. , & ford , h. 2007 , apj , 670 , 928 bouwens , r. j. , illingworth , g. d. , oesch , p. a. , et al . 2010a , apj , 709 , l133 bouwens , r. j. , illingworth , g. d. , oesch , p. a. , et al .
2010b , apj , submitted ( arxiv:1006.4360 ) bruzual , g. , & charlot , s. 2003 , mnras , 344 , 1000 bunker , a. j. , wilkins , s. , ellis , r. , et al .
2010 , mnras , in press ( arxiv:0909.2255 ) coleman , g. d. , wu , c .- c . , & weedman , d. w. 1980 , apjs , 43 , 393 cowie , l. l. , songaila , a. , hu , e. m. , & cohen , j. g. 1996 , aj , 112 , 839 dickinson , m. , stern , d. , giavalisco , m. , et al .
2004 , apj , 600 , l99 giavalisco , m. , ferguson , h. c. , koekemoer , a. m. , et al .
2004a , apj , 600 , l93 giavalisco , m. , dickinson , m. , ferguson , h. c. , et al .
2004b , apj , 600 , l103 grazian , a. , fontana , a. , de santis , c. , et al .
2006 , a&a , 449 , 951 guhathakurta , p. , tyson , j. a. , & majewski , s. r. 1990 , apj , 357 , l9 kalirai , j. s. , mackenty , j. , rajan , a. , et al .
2009 , instrument science report wfc3 2009 - 31 kashikawa , n. , shimasaku , k. , yasuda , n. , et al .
2004 , pasj , 56 , 1011 kinney , a. l. , calzetti , d. , bohlin , r. c. , et al .
1996 , apj , 467 , 38 koekemoer , a. m. , fruchter , a. s. , hook , r. n. , & hack , w. 2002 , the 2002 _ hst _ calibration workshop , ed . s. arribas , a. koekemoer , and b. whitmore ( baltimore : stsci ) , 337 komatsu , e. , dunkley , j. , nolta , m. r. , et al .
2009 , apjs , 180 , 330 labb , i. , gonzlez , v. , bouwens , r. j. , et al .
2010 , apj , 708 , l26 ly , c. , malkan , m. a. , treu , t. , et al . 2009 , apj , 697 , 1410 nonino , m. , dickinson , m. , rosati , p. , et al .
2009 , apjs , 183 , 244 madau , p. 1995
, apj , 441 , 18 oesch , p. a. , bouwens , r. j. , illingworth , g. d. , et al .
2010 , apj , 709 , l16 oke , j. b. , & gunn , j. e. 1983 , apj , 266 , 713 pickles , a. j. 1998 , pasp , 110 , 863 rafelski , m. , wolfe , a. m. , cooke , j. , et al . 2009 , apj , 703 , 2033 reddy , n. a. , steidel , c. , pettini , m. , et al .
2008 , apjs , 175 , 48 reddy , n. a. , & steidel , c. c. 2009 , apj , 692 , 778 ryan , r. e. , hathi , n. p. , cohen , s. h. , et al .
2007 , apj , 668 , 839 santini , p. , fontana , a. , grazian , a. , et al .
2009 , a&a , 504 , 751 sawicki , m. , & thompson , d. 2006 , apj , 642 , 653 schechter , p. 1976 ,
apj , 203 , 297 stark , d. p. , ellis , r. s. , chiu , k. , et al .
2010 , mnras , submitted ( arxiv:1003.5244 ) steidel , c. c. , giavalisco , m. , pettini , m. , et al .
1996 , apj , 462 , l17 steidel , c. c. , adelberger , k. l. , giavalisco , m. , et al .
1999 , apj , 519 , 1 steidel , c. c. , adelberger , k. l. , shapley , a. e. , et al .
2003 , apj , 592 , 728 steidel , c. c. , shapley , a. , pettini , m. , et al .
2004 , apj , 604 , 534 vanzella , e. , cristiani , s. , dickinson , m. , et al .
2008 , a&a , 478 , 83 windhorst , r. a. , cohen , s. h. , hathi , n. p. , et al .
2010 , apjs , submitted ( arxiv:1005.2776 ) wuyts , s. , labb , i. , schreiber , n. m. f. , et al .
2008 , apj , 682 , 985 yan , h. , windhorst , r. a. , hathi , n. p. , et al .
2010 , raa , in press ( arxiv : 0910.0077 ) mag ( solid black lines ) , expected colors of stars ( black dots ) from @xcite , and tracks of low - redshift ellipticals ( gray lines ) from @xcite and @xcite . though stars clearly land in the selection region ,
these are easily removed by simple morphological criteria , as in @xcite .
[ bottom - left ] shows the color - color plot with the @xmath16-dropout selection region , [ bottom - middle ] shows the selection of the @xmath17-dropouts and [ bottom - right ] shows the selection of the @xmath18-dropouts .
the gray data points in the bottom panels are _ all _ objects in the catalog .
average uncertainties in the color measurements are shown as the error bar in the lower right corner .
red points indicate the selected dropouts , while gray data points in the selected region were excluded by other criterion as given in [ sample ] . ]
-dropout ( @xmath20 @xmath29 24.5 mag ) with a spectroscopic redshift of @xmath94 , the object in the middle is a @xmath17-dropout ( @xmath20 @xmath29 23.9 mag ) with @xmath95 , and the object at the bottom is a @xmath18-dropout ( @xmath20 @xmath29 24.7 mag ) with @xmath96 .
each stamp is 3 on a side , has north up , and has a pixel scale of 0.09/pixel . ]
shows distribution of 55 ( 22 ) @xmath16-dropouts with photometric ( spectroscopic ) redshifts .
the average redshifts are @[email protected] and @[email protected] , [ middle ] shows the distribution of 117 ( 57 ) @xmath17-dropouts with photometric ( spectroscopic ) redshifts .
the average redshifts are @[email protected] and @[email protected] , and [ right ] shows the distribution of 203 ( 52 ) @xmath18-dropouts with photometric ( spectroscopic ) redshifts .
the average redshifts are @[email protected] and @[email protected] . ] .
the best fit schechter function parameters are shown for each lf .
the gray shaded region shows uncertainty in the lf based on 1-@xmath31 uncertainties in @xmath10 and @xmath9 . in the leftmost panel , @xmath10 is kept fixed while fitting the schechter function , so the uncertainty indicated with the gray shaded region is based on 1-@xmath31 uncertainty of @xmath9 only .
the more uncertain brightest point in the leftmost panel occurs because of very limited statistics and does not contribute to the best fit parameters , while the uncertain faintest points in the middle and the rightmost panel are at the limit of our observations , and could also be affected by low redshift interlopers .
the vertical error bars in our data are 1@xmath31 poisson uncertainties . ]
@xmath16 & 66 & 22 & 55 + & & [email protected] & [email protected] + @xmath17 & 151 & 57 & 117 + & & [email protected] & [email protected] + @xmath18 & 256 & 52 & 203 + & & [email protected] & [email protected] + @xmath16 & [email protected] & [email protected] & 1.27 ( fixed ) + @xmath17 & [email protected] & [email protected] & [email protected] + @xmath18 & [email protected] & [email protected] & [email protected] + | we combine new high sensitivity ultraviolet ( uv ) imaging from the wide field camera 3 ( wfc3 ) on the hubble space telescope ( hst ) with existing deep hst / advanced camera for surveys ( acs ) optical images from the great observatories origins deep survey ( goods ) program to identify uv - dropouts , which are lyman break galaxy ( lbg ) candidates at @xmath0 .
these new hst / wfc3 observations were taken over 50 arcmin@xmath1 in the goods - south field as a part of the early release science program .
the uniqueness of these new uv data is that they are observed in 3 uv / optical ( wfc3 uvis ) channel filters ( f225w , f275w and f336w ) , which allows us to identify three different sets of uv - dropout samples .
we apply lyman break dropout selection criteria to identify f225w- , f275w- and f336w - dropouts , which are @xmath2 , @xmath3 and @xmath4 lbg candidates , respectively .
we use multi - wavelength imaging combined with available spectroscopic and photometric redshifts to carefully access the validity of our uv - dropout candidates .
our results are as follows : ( 1 ) these wfc3 uvis filters are very reliable in selecting lbgs with @xmath5 , which helps to reduce the gap between the well studied @xmath6 and @xmath7 regimes ; ( 2 ) the combined number counts with average redshift @xmath8 agrees very well with the observed change in the surface densities as a function of redshift when compared with the higher redshift lbg samples ; and ( 3 ) the best - fit schechter function parameters from the rest - frame uv luminosity functions at three different redshifts fit very well with the evolutionary trend of the characteristic absolute magnitude , @xmath9 , and the faint - end slope , @xmath10 , as a function of redshift .
this is the first study to illustrate the usefulness of the wfc3 uvis channel observations to select @xmath11 lbgs .
the addition of the new wfc3 on the hst has made it possible to uniformly select lbgs from @xmath12 to @xmath13 , and significantly enhance our understanding of these galaxies using hst sensitivity and resolution . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.